Title Page
 Table of Contents
 List of Tables
 List of Figures
 Section I: A survey to determine...
 Section II: Mechanisms in cx. nigripalpus...
 Biographical sketch

Title: Mechanisms in mosquitoes responsible for variation in susceptibility to infection by Dirofilaria immitis (Leidy), etiologic agent of canine heartworm disease /
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00099247/00001
 Material Information
Title: Mechanisms in mosquitoes responsible for variation in susceptibility to infection by Dirofilaria immitis (Leidy), etiologic agent of canine heartworm disease /
Physical Description: xi, 168 leaves : ill. ; 28 cm.
Language: English
Creator: Sauerman, Donald M., 1940-
Publication Date: 1980
Copyright Date: 1980
Subject: Mosquitoes as carriers of disease   ( lcsh )
Dirofilaria immitis   ( lcsh )
Canine heartworm disease   ( lcsh )
Entomology and Nematology thesis Ph. D
Dissertations, Academic -- Entomology and Nematology -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis (Ph. D.)--University of Florida, 1980.
Bibliography: Bibliography: leaves 157-167.
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Donald M. Sauerman Jr.
 Record Information
Bibliographic ID: UF00099247
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000014512
oclc - 07216098
notis - AAB7739


This item has the following downloads:

PDF ( 10 MBs ) ( PDF )

Table of Contents
    Title Page
        Page i
        Page ii
    Table of Contents
        Page iii
        Page iv
    List of Tables
        Page v
        Page vi
        Page vii
    List of Figures
        Page viii
        Page ix
        Page x
        Page xi
        Page 1
        Page 2
        Page 3
        Page 4
    Section I: A survey to determine the natural potential vectors of canine heartworm disease in Vero Beach, Florida
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
    Section II: Mechanisms in cx. nigripalpus and ae. aegypti responsible for variable patterns of susceptibility to infection by d, immitis
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
    Biographical sketch
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
Full Text




FOR TIE D-Gi:tS ul' iP X';t CF i JOI




I would like to thank Dr. Jai K. Nayar, for the use of his

laboratory and facilities and for his patience, guidance and support.

I would also like to thank Dr. Richard Bradley, who was instru-

mental in getting this investigation started, and Dr. Jerry Butler,

who provided constant help in a variety of matters along the way.

The support and patience of my wife and children were constant

and boundless.



LIST OF TABLES................................................v

LIST OF FIGURES............................................viii

ABSTRACT..................................................... ix


General Rationale..............................1

The Problem.....................................

BEACH, FLORIDA.... ............ .................... 5


Materials and Methods.........................12

Results and Discussion........................20



Literature Review............................50

Definitions and Usage.........................43

Materials and Methods.........................46

Additional Procedures and Materials............66

Data Presentation and Statistical

Experimental Results and Discussion............74

Culex Susceptibility Patterns -
Background................................ 74

The Effects of Hindrance of Migra-
tion of Microfilariae to the Mal-
pighian Tubules ...........................75

The Relationship of Hemolysis and
Crystal Formation to Susceptibil-
ity of Cy. nigripalpus to Infection
by D. immitie .............................87

Evaluation of Alterations in Sus-
ceptibility Level of Cx. niripal-
pus to D. immitis infection induced
by Additives in the Blood Meal............95

The Determination of ;ources of
Microfilarial Vitality Loss..............101

Mode of Action of the aEcco-
Pharyngeal Armature.......................108

Conclusions Ox. -ld ii 1. 1........... 118

Ae. aegyptiV Susceptibility Patterns -
Background..... ........................ ...122

The Effects of Hindrance Movement
on Susceptibility of Ae. aegypti to
D. immitis Infection .................... 122

The Effect of Atreptic (Nutritional)

The Effect of Antiblastic Factors........124

Hemolymph Factors ......................127

The Chronology of Arrest of Development..128

Conclusions Ae. ae*ypti................130


BIOGRAPHICAL SKETCH................................................. 168


Table Page

I-1. Natural Potential Vectors of Canine Heartworm
Disease in Vero Beach, Florida. Combined (55)
Collections by Species, Trap Site and Trap Type....... 152

1-2. Natural Potential Vectors of Canine Heartworm
Disease in Vero Beach, Florida. Species Collected
and Species Found Positive for Dirofilaria immitis
Infection............................................ 155

1-3. Natural Potential Vectors of Canine Heartworm
Disease in Vero Beach, Florida. Analysis of 16
Positive Females From 5 Species...................... 134

II-1.. Host Efficiency (Portion of Prelarvae in Tubules
of Total Ingested Microfilariae) and Relationship
Between Number Ingested and Prelarvae Developing
at +24 Post Feed in Culex nigripalpus............... 135

11-2. The Effects of Hindrance of Microfilarial Movement
On Susceptibility of Cx. nigripalp s to D. immitis
Infection Via "Normal" Bloodfeeding Mode, Measured
by Mean (5) Prelarvae Per Female With Confidence
Interval (CI) at P=0.05.............................. 136

II-5. Comparison of Coagulation Patterns in Bloodmeals
from Groups of Cx. nigripalpus Fed Directly On a
Dog and Fed Via Membrane On Both Anticoagulant-
Treated Dog Blood (AC) and Glass Bead-Defibrinated
Dog Blood (GBDF), Tested by Coverslip Compression
For Clot Presence.............................. 13... 7

11-4. Coagulation Times (Mean Minutes + Confidence Interval
(CI) at P=0.05) of 50 )l Samples of Untreated Dog
Blood Added to 5 ul 0.90 NaCl Containing Dissolved
Extract of Cx. niFrialpus Salivary Glands............ 138

II-5. The Effect of Midgut Extracts On Coagulation Time
of 50 pl of Dog Blood .................................159

11-6. In Vivo Hemolysis of Chicken fSrythrocytes in
Mosquitoes Fed Directly on the Host and Via
Membrane on Anticoagulant-Treated or Glass Bead-
Defibrinated (GBDF) Blood............................ 141

1-7. Hemolysis of Chicken Erythrocytes Following In
Vivo, Non-Per Os Introduction of Cell Ulnpensions
and In Vitro Testc on Extracts....................... 142

11-8. Crop Ccntents Acidity: Distribution in Eosquitoes
and Effects on D. immitis Microfilariae and Chick
Red Cells........................................... 143

II-9. The Effects of Alteration of the Blood Food Chemical
Composition On Susceptibility of Cx. nigipalpus to
D. immitis Infection. Mean (X) Prelarvae Per Mosquito
and t 0.05 Confidence Interval (CI) and Susceptibility
Profile ...............................................144

II-10,. he Effects of Alteration of Maintenance Sugar and
Blood Food Chemical Composition On Susceptibility of
Cx. nigripalpus to D. immitis Infection. Mean (X)
Prelarvae Per Mosquito and t 0.05 Confidence Interval
(CI) and Susceptibility Profile.......................146

II-11. Microfilarial Vitality; Tests for Antiblastic Factors
in Cx. ni Rilpus; Mean (X) Microfilariae + Standard
Error (SE) With Distribution According to Precent

11-12. Microfilarial Wound Assay............................. 49

II-13. D. immitis Microfilariae Sucrose Tolerance Test....... 50

11-14. Hemolysis of Chicken Erythrocytes; Ingestion of
Chicken Erythrocytes Via Sugar-Feeding Mode (BF-SF)
Mean (X) With Standard Error (SE) of % Hemolysis
Per Female and Hemolysis Profile .......... ........ 51

11-15. Microfilarial Vitality Patterns Microfilarial
Morbidity in Mosquitoes Infected Via Abnormal (Sugar
Feeding) Mode..................................................52




II-16. Variation in Susceptibility Levels of he. aegypti
to '1. in.,itis Infection. After Various Treatments
to3upply Atreptic Factors or Delete Antiblastic
Factors................................................... 155

11-17. Distribution of Mosquitoes According to Stages of
D. immitis. Harbored 120h After Being Injected
Into the Hemocoel, and Number and % Moribund of
Larval Types and Number of Host Responses.................155

II-18. Early Chronology of Arrest of Development of
D. immitis Prelarvae in the Malpighian Tubules
of Ae. aegynti Fed Directly on the Infectious
Dog, With An. quadrimaculatus Procedural Controls
(PC), as Measured by Portion of Normal and Abnormal
Prelarvae With the Passing of Time. Mean (x) +
Standard Error (SE) With Percent (%).....................156




I-1. Map of Vero Beach, Florida, showing survey trap site...... 15

1-2. Modified, high-survival, CDC-light trap receptacle........ 17

II-1. Feeding methods: direct feeding of mosquitoes on legs
of dog inserted through access sock ....................... 48

11-2. Feeding methods: direct feeding of mosquitoes on shoulder
of dog laid down on top of cage........................... 48

11-3. Feeding methods: membrane-feeder, schematic............... 50

11-4. Feeding methpds: membrane feeder in position on top
of cage, with water bath reservoir (left) containing
pump and thermoregulator.................................. 51

11-5. Malpighian tubule squash for prelarval count.............. 57

11-6. Malpighian tubule squash for sausages .................... 57

11-7. Special apparatus for blood feeding mosquitoes in
the sugar feeding mode-schematic.......................... 68

11-8. Serial sections of a) anterior midgut of Cx. nigrip-
alpus and b) posterior midgut of Cx. nigripalpus,
showing free nuclei from hemolyzed cells, and c)
posterior midgut of Ae. aegypti, showing aggulination
but no hemocolysis....................................... 91

11-9. Sections trough buccal cavity showing dorsal pharyn-
geal valve,(bucco-pharyngeal armature)....................115

II-10. Distribution of wounds in microfilariae from Cx. nig-
ripalpus.. ... ......... ...............................106

II-11. Melanization host response of Ae, aegypti to D. iummitis
in the malpighian tubule ....................*.............126

11-12. Hemacyte host response of Ae. aegypti to D. immitis
in the hemocoel..........................................129

Abstract of Dissertation Presented to the
Graduate Council of the University of Florida
in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy


Donald M. Sauerman Jr.
December, 1980

Chairman: Jerry Butler
Co-chairman: R. E. Bradley
Major department: Entomology and Nematology

Based on the hypothesis that the susceptibility level of a potential

mosquito vector to filariid nematode infection is determined by the presence

or absence of physiological or morphological factors that manifest

themselves as barriers at either the midgut or specific target tissue

level, the patterns of susceptibility and mechanisms responsible for these

patterns were investigated by employing an addition: deletion strategy

for experimental delineation of mechanisms, using model mosquitoes to

represent 3 frequently encountered susceptibility categories. The

mosquitoes, Culex nigripalpus for partial susceptibility determined by

a gut barrier, Aedes aegypti for partial susceptibility determined by a

specific target tissue barrier, and Anopheles quadrimaculatus for

susceptible control, were chosen for their ability to sharply display

the attributes of their particular pattern.

Neither the variability of particular patterns nor the mechanisms

that generate them have been adequately explained or, with few exceptions,

thoroughly studied. It was the aim of the research reported in this

thesis to more clearly characterize these factors, since the variability

of the patterns suggest that this might be an aspect of the vector-parasite

relationship that is vulnerable to manipulation by man. A preliminary

survey for natural potential vectors implicated 5 species, Aedes

taeniorhynchus, Culex ni&gipalpus and Culex quinquefasciatus, as most

likely transmitting the disease in the Vero Beach, Florida, area.

Susceptibility level was determined in Cx. nigripalpus to be a

function of 2 factors, hindrance of microfilarial movement through the

midgut by rapid clotting of the bloodmeal and wounding of microfilariae;

the number of microfilariae ingested by Cx. nigripalpus fed on a highly

microfilaremic (approximately 75 microfilariae/ul blood) dog was neither

a factor in the determination of nor could be useful as an estimator of

the susceptibility level. The addition of anticoagulants to blood fed

to Cx. nipripalpus via a membrane feeder failed to alter the susceptibility

pattern, probably because they failed to prevent coagulation. However,

the absence of salivary anticoagulant was not a factor, as extracts of

salivary glands prevented clotting of venous blood. The presence of a

powerful coagulant in midgut extracts was demonstrated, and, together

with the production of shifts in the susceptibility pattern produced

when Cx. nigripalpus were fed on glass bead-defibrinated blood, support

the hypothesis that hindrance of movement of microfilariae is at least

a partial determinant of susceptibility in the species.

Hemolysis and crystal formation were associated with, but not

responsible for, microfilarial mortality in the Cx. nigripalpus midgut.

Non-wounded but moribund microfilariae found in crops were probably the

result of flora-produced low crop contents pH. Wounding, which occurred

only in Cx. nigripalpus fed per os or sugar mode with plasma added, was

demonstrated to be the other major determinant of susceptibility level,

with the rabidly oscillating bucco-pharyngeal armature being the most

likely source of the wounds. The involvement of the food-routing switch

mechanism seemed likely, based on wound pattern variation associated

with feeding mode, and explains shifts in susceptibility produced in

mosquitoes fed on buffer-treated blood.

Conventional addition: deletion strategies were non-productive in

revealing the mechanism in Aedes aegpti that produces arrest of develop-

ment of prelarvae in the tubules, and hindrance of movement was not a

factor. While the melanization inhibitor, reduced glutathione, has

some effect on arrest of development, only interval-feeding significantly

altered the pattern. The phenomenon starts to occur within 6 hours of

ingestion and has a chronological aspect. The intracellular-station

hypothesis is not supported by the finding here of developing forms in

the hemocoel after injection of microfilariae.


General Rationale

The fundamental rationale for the research in this dissertation

is to support man's effort to control filariasis in his own kind and

in his animals. Freedom from disease is requisite for the efficient

use of our resources, which can translate into issues as basic as the

success of birth control programs in populations that are breeding

rapidly simply because the chance for any one child's survival is so

small. The breaking of the epidemiological triad can be attained by

assaults on any of its links, and often involves a variety of approaches,

as with filariasis, for example, which is attacked with assaults on the

vector with insecticides and treatment of hosts with paraciticides.

Treatment of the definitive host, either prophylactically or thera-

peutically, has certain advantages and limitations depending on the

particular disease, but is definitely in ascention at the present time

as one of man's major strategies against disease. In many diseases,

the vector is placed under assault by man using the weapons of pest

control, such as pesticides, insect growth regulators, genetic methods

(sterile release; gene insertion), and integrated pest management.

As with treatment of the definitive host, this strategy's weapons have

their advantages and limitations. Our traditional weapons, the

pesticides, have been shown by our own scientific community to be more

accurately labelled biocidess," targetless weapons that have had a


way of coming back to man full-circle. However, judicious use of

scientifically established levels of biocides is still one of the

components of modern pest management weaponry. Growth regulators and

genetic methods are likewise limited to particular situations. Of

course, the development of these techniques is on-going, with constant

pressure for expansion of application for each system. Integrated

pest management is likewise only just surfacing as a utilized concept,

and awaits only the refinement of its ecological data acquisition

strategies to expand its presently limited applicability. And,

setbacks in systems, such as development of resistance or unacceptable

contamination levels, add to the total cost of a system already

expensive because of high development cost, itself the result of

labelling procedures deemed necessary to prevent further contamination

of the environment or the introduction of unseen health hazards.

The satisfying of man's need as viewed from the standpoint of the

public health researcher means solving epidemiological problems. Basic

research serves as the breeding ground of applied technologies, a

relationship that is fundamental in the generation of solutions since

a thorough knowledge of a disease usually precedes its control. In

basic epidemiological research, the concept of the triad is often

utilized to conceptualize the interrelatedness of elements. The

arrows connecting parasite-to-vector, vector-to-host and host-to-

parasite imply a complexity of elements, since the arrows travel in

both directions; for example, vector-parasite relationships involve

both effects of the vector on the parasite and effects of parasite on

vector, which is logical, since the vector is a host.

One aspect of the epidemiological problem in the mosquito-

onchocercid relationship that could prove to be of profound importance

to the attainment of the thorough knowledge needed to conquer these

diseases is the variation in susceptibility of potential vectors to

the disease. The mechanisms that allow one mosquito to carry huge

filariid larval loads while rendering others totally refractory have

just begun to be investigated, with the first research specifically

aimed at resolving this problem being less than 30 years old. It is

the purpose of the research reported in this dissertation to increase

our knowledge of these mechanisms.

The Problem

The susceptibility to filariid diseases in mosquitoes, while

variable even at the individual level in some species, and from

strain to strain in others, nevertheless fallsinto a limited number

of generally valid categories, based on success in supporting larval

development and transmitting the infectious stage. Most mosquitoes are

either susceptible, partially susceptible or refractory as inter-

mediate hosts, indicating that host specificity, perhaps the most

vulnerable link in the parasite's life-cycle for man to exploit,

is involved in the production of these different susceptibility levels;

characteristic patterns of susceptibility fall into as few as 6 or 7


The physiological mechanisms responsible for generation of these

groups is poorly understood in most cases, although much has been

learned. Two major barriers to successful development of parasites

have been described in general terms for filariids, the gut barrier

and specific host tissue barrier, although, in general, only one

type of barrier exists in any one group.

For this study, mosquitoes were chosen on the basis of their

suitability as representatives of a particular type of susceptibility

pattern, and this, in turn, usually resolves itself, physiologically,

into a confrontation with one of the barrier-types mentioned above.

The models possessed clear and sharp expressions of the patterns

induced by the mechanisms, increasing the chance of detecting experi-

mentally-induced dislocations in the vector-parasite relationship.

This, in turn, will lead to the characterization of the underlying

mechanisms which are the genetically-determined transducers of suscep-

tibility-level determinants.

Prior to initiating laboratory experimentation, a survey for

potential vectors of canine heartworm disease in the study area was

conducted. The survey provided "hands-on" experience with the prob-

lems inherent in such epidemiological studies, as well as exposure

to the business-end of disease transmission. The result of the

survey and its relevance to susceptibility studies is discussed in

the first section, followed by the laboratory studies.




Since 1900, when together Grassi and Noe first implicated moe-

quitoes as intermediate hosts of Dirofilaria immitis, the etiologic

agent of dirofilariasis, or canine heartworm disease, a succession of

researchers have sought to establish which species are actually in-

volved in transmitting the disease. With few definite exceptions,

canine heartworm disease is nearly as widely distributed throughout

the world as its primary definitive host, while very few mosquito

species are that cosmopolitan, exploiting, as they do, but also re-

stricted to, a variety of habitats that may or may not bring them

into the epidemiological web. Therefore, the determining of vector

species when the disease occurs can be a local or geographic regional

problem. The disease is characterized by its emanation from nidi of

variable size, from neighborhoods to entire cities, and the intensity

of the disease, measured by incidence in dogs or wild mosquitoes, is

a function of the optimizing of the requirements of the epidemiological

triad, which are rigorous. The infectious dog must occupy the same

time and space as the bloodseeking and susceptible mosquito. If the

mosquito is to vector the parasite, it must first survive the infection

itself and then survive long enough (about 2 weeks) to allow parasite

development, after which a second or third bloodmeal taken on a dog or


other suitable definitive host completes the cycle. That canine

heartworm disease is endemic on the oouth Atlanti. tad Gulf coasts

of the United States attests to a confluence of satisfied epidemiological

requirements, including host and vector density and the surmounting of

array of further biological barriers. But nidi have been developing

throughout the country (Harrison et al., 1965; Otto, 1972), probably as

a result of human population mobility, with the result that formerly

"clean" areas are now diseased; one often-cited example is California,

where the first autochthonous cases were reported in 1970 (McGreevy

et al., 1970).

One of the basic requirements for vectorship is susceptibility

to the disease. That mosquitoes vary in susceptibility to D. immitis

was first reported by Feng (1930) in leiping, China, and this was

quickly followed by the investigation of HTu (1951) in the United

States, who listed 12 species in 5 genera from 4 regions as known

susceptible hosts for D. immitis, while determining the susceptibility

of 8 species himself. Hu concluded that a {god host species is one

which shows a high group percentage of infection and high individual

intensity of infection; he addressed susceptibility variation directly

and was one of the first investigators to comment on individual variation

within a species, an important milestone. In the same decade, Roubaud

(1937) reported geographical strain variation in Aedes aegypti's

susceptibility to D. immitis infection. Further reports on potential

vectors of CHD came from as far as the Phillipines (Rosario, 1956);

and Yen's (1938) basic study was the first to establish clear-cut

categories into which most mosquitoes could be placed based on their

susceptibility. The physiological basis of susceptibility variation

was finally addressed by Kartman (1955a), in a classic investigation

attempting to link susceptibility with hindrance of migration and

effects from digestive secretions. The determination of experimental

vector potential has continued to be a basic method used to estimate

the importance of a species in a particular area in transmission of

CHD. These investigations include those conducted by Bradley (1953),

Symes (1960), Beam (1965), Bemrick and Moorhouse (1968), Chellappah and

Chellappah (1968), Weiner and Bradley (1970), Suenaga (1973, 1975, 1978),

Intermill (1975), Weinmann and Garcia (1974), Seeley and Bickley (1974),

Nayar and Sauerman (1975), Jankowski and Bickley (1976), Christensen

(1977), and Rogers and Newson (1979). As early as 1970, over 60

species, worldwide, had been proven to be "susceptible" to D. immitis

infection (Ludlam et al., 1970). The significant feature which emanates

from these studies taken collectively is that to a greater or lesser

degree, and with varying patterns, a large number of mosquito species

are susceptible to D. immitis and have therefore satisfied a basic

requirement for vectorship.

In conferring potential vector status, susceptibility determination

can be definitive without further investigation if a particular species

is solitary in the diseased area, and if increased rates due to immi-

gration of previously infected dogs can be ruled out. However, there

are more commonly several candidate species that qualify as far as

susceptibility alone is concerned, and additional information must be

acquired before ranking the species regarding relative vector potential.

As Rosen (1954) points out, "In nature, the importance of a given species

as a vector may be influenced by such factors as its abundance, the

frequency with which it feeds on vertebrate hosts of the parasite, and

its susceptibility to adverse effects of the developing filarial larvae.

These factors may be of equal or greater importance than its relative

susceptibility to infection" p. 325. Ecology and biology are important

factors (Ludlam et al., 1970), and these include breeding sites, molts

per year, longevity, flight range and, certainly, strength of host

preference. There is fairly universal agreement that there is not

enough information on natural infection and transmission (Ludlam et al.,

1970; Christensen and Andrews, 1976). Ideally, such studies would

require a high level of enzooticity, enabling the capture of enough

infected mosquitoes to allow a "differential" count to be made. Further-

more, experimentally determined susceptibility levels may differ

radically from those measured in field populations, implying the

presence of barriers in nature that do not manifest themselves in the

laboratory. Thus, noting that only 1 of 466 field collected Anopheles

punctipennis harbored developing forms while none had infective stages,

both patterns radically different from experimentally-determined

susceptibility levels for that species, Christensen and Andrews (1976)

concluded, "It is possible that the parasite is unable to develop to

the infective stage under natural conditions or may cause a high degree

of mortality with this particular strain of Anopheles punctipennis" p.279.

Furthermore, species are subject to elimination from consideration as

vectors because of obvious mitigating factors, such as absence of geo-

graphic sympatricity, chronological confluence, rigid host preferences

(rarely), and other often less obvious factors barring the mosquito

from candidacy. Such factors, as host preference, are commonly used to

explain absence of infection in a particular species (Christensen and

Andrews (1976) with Culex tarsalis, for example), but host preference

can be a matter of host availability or population density (Edman and Taylor

1968), and eliminating from candidacy a species because it is rural

is risky since dispersal or migration can suddenly bring it into

contact with population centers and the epidemiology of the disease.

Likewise, while number of molts can be a factor, it is probably less

important than longevity and ecological focusing, important factors

in establishing a restricted nidus with species such as Ae. sierrensis

(Weinmann and Garcia, 1974), while a short-lived mosquito, regardless

of the extent of multivoltinism, is a poor candidate. Longevity is

also altered by the effects of the parasite on the host, a phenomenon

addressed by several investigators, and mortality in infected mosquito

populations can soar above 9Q0 long before parasite development is

complete (Weiner and Bradley, 1970).

The isolation of infected mosquitoes in nature obviates the

acquisition of much circumstantial evidence. Confluence of host and

vector has occurred without the artifaces of the laboratory. If there

are developing forms, she is susceptible without further ado, and if

L3 infective stages are found in the head and proboscis, her candidacy

as a potential vector, and thereby that of her species, is greatly

strengthened. Host preference studies are made moot, and the ranking of

the species in the population of all mosquitoes is possible. The

only step untaken is proof of ability to actually experimentally

transmit from dog-to-dog, which has been done, but only for a compara-

tively few species, namely, Ae. togoi (Kume and Itagaki, 1955), An.

quadrimaculatus (Newton, 1957), As. aegypti (Taylor, 1960), Ae. vigilax

(Bemrick and Moorhouse, 1968) and Ae. canadeniL (Bickley et al., 1977).

However, transmission studies, as experimental susceptibility studies,

do not demonstrate the natural act, only the potential for its occur-


The "paucity" of data on natural infection rates in mosquitoes,

and the need to acquire it, has been addressed by Ludlam et al.,

(1970), and they feel that it is attributable to the work involved in

collecting, identifying and dissecting numerous specimens to obtain

meaningful rates. With all other things being equal, increases in

infection rates are attributable to an increase in number of infected

dogs or a decrease in the number of diluting non-infected mosquitoes.

Yields of infected mosquitoes from collections made with a variety of

methods are often less than 2.1%; for example, Rosen (1954) collected

and dissected over 4000 mosquitoes, representing 7 species, but only

10 were positive in 2 species. Two species (Ae. aogypti and Cx.

quinquefasciatus) positive for susceptibility in experimental tests,

were refractory. Similarly, a collection of 4746 mosquitoes representing

19 species yielded a meager 3 mosquitoes with let stage larvae and

prelarvae in 2 specimens, all from one species (Ae. vexans) (Bemrick

and Sandholm, 1966) and of 2312 Cx. quinquefasciatus collected, only

1.45% contained developing heartworm larvae (Villavaso and Steelman,

1970). In field studies supplementing lab studies, engorged-only

mosquitoes oolleoted from a sentry dog complex where nearly one-half

of the animals were suspected of being infected, infective L3 from

individually dissected mosquitoes were found in 15.3% of one species,

0.53% of another and none in 4 other species (Intermill and Frederick,

1970). Suenaga (1973, 1975, 1978), in three separate studies conducted

in different cities, determined natural infection rates ranging from

0.36 to 5.% in 9 mosquito species trapped with human and dog bait

traps or with light traps. Of 542 specimens representing 10 species

collected at a kennel with a record high incidence of diseased dogs,

only 4 infections were found in 5 species (Bickley et al., 1976),

and D. immitis was recovered from 4.6% of Ae. trivittatus, 0.426 of

An. punctipennis (Christensen and Andrews, 1976), 0.03 0.82 in 4

species of 2 genera in Malaysia (Basio and Cheong, 1977). Eleven infected

mosquitoes representing 5 species were found in a collection of 3445

specimens (Arnott and Edman, 1978) and 1.4f of 3294 mosquitoes collected

in Connecticut were positive for D. immitis (Magnarelli, 1978), while

a rate of 0.21% was obtained from determinations of 25 pooled species,

92 D. immitis being found in 3 species from a collection of 45,627

individuals (Lewandowski et al., 1980). Thus, even when conditions

are optomized for finding infected mosquitoes, the rates are low and

ranking by rate therefore difficult, and a major consideration in trap

site or collection area selection must be yield if meaningful results

are to be obtained (Ludlam et al., 1970).

While it is certainly of concern that D. immitia is succeeding in

establishing sylvan reservoirs in wild animals, with microfilariae

having been discovered in the blood of wolverines (Williams and Dade,

1976) and sea lions (Taylor et al., 1961), of greater immediate concern

from the standpoint of human public health is the occurrence of heart-

worm disease as a zoonosis. While the other dirofilariids associated

with man's domestic animals (Dirofilaria repens) or closely impinging

wild animals (Dirofilaria tenuis in the racoon) are of little medical

concern outside of dermatology since they are nearly always found in

subcutaneous tissue, D. inmitis is less well-adapted to man, and after

some development, probably in the right heart, it lodges in lung

arterioles and dies; it is then thrombocized, producing pulmonary

infarctions which, while largely symptomless, can cause a variety of

symptoms from cough to chills to fever, and can, more importantly,

lead to misdiagnosis as early bronchial carcinoma (Dissanaike, 1979),

which is undoubtedly stressing on the patient. Since the 68 cases

reported during the last 2 decades have come from only 3 countries

(USA, Japan, Australia), and 70% of the cases are symptomless, found

by accident during radiography, the tip-of-the-iceberg phenomenon is

probably occurring (Dissanaike, 1979; Weinmann and Garcia, 1974).

Human pulmonary dirofilariasis has been described as a disease

which "affects individuals living near coastal breeding areas" p. 407

(Robinson et al., 1977), and, indeed, most of the cases from the

United States have been reported from Florida. With the immense

increase in canine heartworm disease in Florida, with 493 reported

cases in 1966, 1,250 in 1970, and 5,558 cases in 1976 (Florida

Morbidity Statistics, 1966, 1970, 1976), the determination of natural

potential vectors and their rate of infection constitutes an important

prerequisite to a full understanding of the disease. The purpose of

the present survey is to determine the natural potential vectors of

heartworm in a region of Florida highly endemic for heartworm and

therefore where introduction into humans is undoubtedly commonplace.

Materials and Methods

The basic strategy for the acquisition of data needed to determine

the natural potential vectors of D. immitis is to collect wild mosquitoes

from suspect areas. Conducting studies where there is an "adequate"

source of infectious dogs is requisite (Ludlam et al., 1970). Field

researchers have made use of a variety of devices and procedures to

more or less maximize the yield of infected wild mosquitoes; however,

the aim of the investigation is what in the final analysis dictates

choice of collecting devices and the sites where they will be put to

use. One such survey restricted collections to within 40 yds. of

human habitation (Rosen, 1954), while others have used established

dog yards (Villavaso and Steelman, 1970), kennel walls in kennels

harboring heartworm infected dogs (Intermill and Frederick, 1970),

hunt clubs with infected dogs (Bickley et al., 1976), areas with

reported heartworm disease with breeding areas nearby (Arnott and

Edman, 1978) and sites of recent infection (Lewandowskiet .. 1980).

While most collection sites are intentionally established in residential,

suburban areas, to capitalize on confluent dog and mosquito populations,

2 investigations (Villavaso and Steelman, 1970; Suenaga, 1973) utilized

5 distinct human (and, therefore, dog) habitats, and the results in-

dicated that while residential areas are in general predictably higher

in the number of infective mosquitoes, other factors than dog density,

such as proximity of breeding sites, especially for domestic mosquitoes,

and owner habits regarding leaving of pets outdoors and unprotected,

influenced the yield of infected mosquitoes.

As there is variation in collection site depending on the aims of

the investigator, there is similar variation in the device selected to

sample the suspect population. Biting collections have been used by

some (Rosen, 1954; Arnott and Edman, 1978), aspiration of engorged

females by others (Intermill and Frederick, 1970), while more commonly

used are light traps (Suonaga, 1975, 1975, 1978; Bickley et al., 1976;

Christensen and Andrews, 1976; Arnott and Edman, 1978; Lewandowski

et al., 1980) and bait traps, usually dog-baited (Bemrick and Sandholm,

1966; Villavaso and Steelman, 1970; Suenaga, 1973; Lewandowski et al., 1980).

but sometimes human-baited (Suenaga, 1973). Rosen (1954) collected

from pigs, humans and dogs, while power aspirators were used to collect

resting mosquitoes from the leaf litter (Arnott and Edman, 1978).

Thus, there are no hard-and-fast rules for making epidemiological

surveys, except that the goal is to find the mosquitoes that are in-

fected, and to find them in sufficient numbers to allow their being

ranked by infection rate. In this study, which was conducted from

April to August, 1977, 3 sites (Figure 1) and 2 trap types were

employed. Site 1, in northeast Vero Beach, was in a residential area

closest (1.5 kn) of the 3 sites to the Indian River salt marshes. Site

2 was in the center of the city's residential area, about 2 km further

west than site 1 from the river. Site 3, in southwest Vero Beach, was

about 4 km from the Indian River, but considerably closer to rural,

agricultural and ranch areas, without the insulating barrier of houses

extant at site 2. Thus, while all 3 sites had substantial dog popula-

tions (personal canvass), the proximity to breeding areas, other than

those normally found near homes, such as containers and bird baths, was

considerably lower for site 2, while sites 1 and 2 were considerably

further from salt marsh. Owner habits towards pets were not determined,

but there was no subjectively distinct difference between the areas in

types of housing or lifestyle.

Mosquito populations at each site were sampled using 1 light

trap (CDC-type) and 1 bait-trap (inaccessible chick-baited, lard can


0 I i


Figure I-1. Map of Vero Beach, Florida, showing survey trap sites.
Scale 1cm 1k
S I^ -1

"-' .. . - " ' *I !
. .. O- 'f." .. 'j
I' - I...-L -

.~ :; ",' _. -i _-- '

Figure 1-I. Map of Vero Beach, Florida, showing survey trap sites.
Scale:' 1cm = 0.51m

trap as used by Nayar (1980), both with carbon dioxide adjuvant. Sawn

blocks of dry ice, 5-6 cm on a side, were placed in 2.54 cm thick

styrofoam boxes, made specifically to hold the blocks, tied shut with

string and suspended a few inches above each trap. Traps were suspended

from available vegetation, not more than 1 m from the ground. They

were set out approximately 1h before dusk and collected at 0800h the

following morning. In both traps, captured mosquitoes had access to

10/ sucrose solution, and the collection container for the light traps

was modified to enhance survival (Figurel-2). After returning to the

laboratory, each collection was lightly chloroformed and placed in

high humidity chambers with access to 1Ml sucrose solution until

dissected. Dissection was initiated as soon as all collections had

been removed from the traps, with each sample anesthetized with chloro-

form and separated according to species. Then, one species at a time,

heads were removed to small saline droplets on siliconized slides; as

many as 30 heads/proboscides could be processed individually per slide

by placing a 24 x 50 mm coverslip over the droplets and evenly compres-

sing the heads. The alimentary tract, including Malpighian tubules,

was drawn from the body of the mosquito into a droplet of saline by

grasping subterminal segments with insect forceps and slowly pulling

backwards while holding the rest of the body. The midgut was discarded

if no bloodmeal remnants were observable under a dissecting scope; all

normal-appearing tubules were subjected to a mild coverslip compression

as with heads, allowing individual inspection, but with less pressure,

unless there were larvae visible or there was obvious pathology (swollen

areas, clear areas) in the tubules, in which case more delicate compres-

sion revealed the stages within the tubule without disrupting them.

to light trap

sugar vial hole

nylon cloth


cloth holding ring

I-t--86rm -m I

sugar vial

nylon cloth

'Figure 1-2. Modified, high-survival, CDC-light trap receptacle.

+ --- 1721rjn 1,

Generally, prelarvae and early sausages are not easily detectable

without coverslip compression, as they are intracellular, and the

tubules must be considerably compressed; such compression easily

destroys beyond recognition the later stages, especially stage II.

Midguts with bloodmeals were ruptured in saline and examined for presence

of microfilariae and type of red cell nucleatedd vs. non-nucleated)

if possible.

Traps were set out and collections recovered twice weekly for

nearly 4 months, yielding a total of 35 collections. Since each

day's collection was composed of 6 traps, each individually processed

by species, the time factor was considerable. Collections were dis-

continued when breeding conditions improved, resulting in samples too

large to process individually.

In addition to processing-time problems, another source of concern

regarding the results of such surveys is the identification of nematode

larvae or their source. Thus, there are not only other dirofilariids

in nature, but there are other animals than the dog with D. immitis

infections (Levine, 1968). Since there is almost always possible

contamination from other filariids, even when collecting engorged

mosquitoes from kennel walls (Intermill and Frederick, 1970), that is,

they could easily be seeking their 2nd or 5rd bloodmeal after having

been infected with a different filariid on their 1st or 2nd, a con-

siderable amount of judgment is needed to make a "presumed" identifica-

tion, and this judgment must take into account the role that wild fauna

might be playing. The rationale employed here are not "fool-proof" and

offer a "presumed" identification in the fashion of other workers in the

field. The strength of the rationale varies, depending on the stage

found and part of the mosquito from which it was recovered. Each

stage will be treated separately as it relates to 4 criteria:

1) measurements, 2) outicular wrinkles, 3) caudal (not anal) papillae

and 4) location in the mosquito:

Microfilariae: The microfilaria is herein defined as that form still

in association with the vertebrate blood, be it in the vertebrate

bloodstream or invertebrate midgut bloodmeal; these have no potential

for further development in this association but must attain lodging in

the target tissue, the Malpighian tubules in this case, prior to

initiation of growth and development. The microfilaria is rejected as

D. immitis if it is 1) less than 4.8 p in breadth or 285 P in length,

which are characteristic of Dipetalonema reconditum (Newton and Wright,

1956); or 2) greater than 327 p, as with Dirofilaria striata (Orihel

and Ash, 1964), or greater than 370 p, as with D. tennis (Pistey, 1958);

3) or if it has cuticular wrinkles or striations, as in D. tenuis

(Pistey, 1958).

Prelarvae: The prelarvae is essentially identical to the microfilariae

except that it is now in the specific target tissue, the Malpighian

tubules. Only dirofilariids use this site, and thus prelarvae of

Dipetalonema are automatically eliminated from consideration while

D. tenuis and D. striata (highly improbable anyway, if the survey by

Orihel and Ash, 1964, is taken into consideration) are delineated as

for MF.

Developing Forms: Absence of information that would allow precise

identification of these forms again precludes precise identification,

but the only probable contaminant is D. tenuis, which has characteristic

cuticular wrinkles, especially prominent in these stages, that allow

it to be differentiated from D. immitis (Pistey, 1958).

L3: when found in the tubules, only D. tenuis is a probable candidate

other than D. immitis, which is rejected if cuticular wrinkles are

present. When found in the head:proboscis, D. immitie is rejected if

the length is greater than 1250 p, most likely Setaria (Becklund and

Walker, 1969), or width less than 17 ), (possibly Foleyella spp.)

(Benach and Crans, 1975; Kotcher, 1941), or greater than 28 p, possibly

Dipetalonema. Absence of caudal papillae eliminates all Foleyella and

possibly D. tenuis, while Dipetalonema reconditum has a long tail with

2 subterminal, ventrolateral ears, much more prominent than the nearly

invisible nubs on D. immitis. It should be noted that Orihel (1959)

was unable to delineate D. tenuis from D. immitis; therefore, the criteria

used here in that regard, especially for L3, are not outside of debate.

Furthermore, rejection of D. striata from the bobcat, based on a

study by Orihel and Ash (1964) who found no D. striata in any of 15

cats dissected, is not without risk, although it should be pointed out

that the cats examined were from central and south Florida. Dipetalonema

arbuta from rodents is less than 18 u in breadth and has 3 caudal papillae.

Results and Discussion

Thirty-five collections of mosquitoes, each comprised of 3 light

traps and 3 bait traps, a total of 210 trap samples, yielded 1030

specimens representing 15 species in 5 genera (Table I-1). Three

species, Cx. nigripalpus, Ae. taeniorhynhus, and Cx. Quinquefasciatus,

accounted for 76' of the total collected, with no other single species

constituting more than 5%. While comparable numbers of Cx. nigripalpus

and Ae. taeniorhynchus were collected by both types of traps, bait

traps accounted for a vast majority of Cx. quinquefasciatus. However,

possibly due to the use of CO2 adjuvant with both systems, the contri-

butions made to the total collection by both types of traps were

comparable (46/ for light traps and 54%o for bait traps). As might

be expected, based on its proximity to the Indian River salt marsh

area, site 1 collections accounted for nearly twice as much (56) of

the Ae. taeniorhynchus collected than the next largest site, site

3 (30%). Similarly, site 3, closest to the rural ranch-grove areas,

accounted for 60h of all Ux. nigripalpus collected. However, the sizable

collections of Ae. taeniorhynchus at site 3 (79/264=30Go) and of Cx.

nigripalus at site 1 (87/405=224) attest to the dispersal or migratory

powers of these species. The large portion (122/200=6C/o) of Cx.

quinquefasciatus at site 3 is likely a reflection of its proximity to

residential sewer treatment systems, not found as a rule at the other

2 sites, which are serviced by a central plant in the city system. The

efficiency for the capture of each species at each site was, with some

exceptions, similar to the rankings between sites, but slightly more than

half of all mosquitoes collected were contributed by site 3 (526/1030),

and only 16)o (165/1030) by site 2, which contributed the fewest total

mosquitoes of the 3 major species collected.

The same 3 species that accounted for 767l of the total collected

accounted for 100% of 16 nematode-parasitized mosquitoes (Table 1-2 ),

2 of which (Cx. quinquefasciatua 2 and Cx. nigripalpus 3) were not

infected but had ingested microfilariae in their bloodmeals, and one

infected mosquito (Ae. taeniorhynchus) with a non-identified L3 (Table

I- 3 ). Thirteen presumed-D. immitis infected mosquitoes, or 1.3A of

all collected (1030), bore a total of 101 larvae of all 4 stages,

one mosquito (Ae. taeniorhynchus 2) harboring both L1 and L2. Only

one mosquito (Cx. quinquefasciatus 3) was found with prelarvae,

indicating since they were active, that it had fed during the previous

24-36 hours. The portions of other developing forms (LI and L2) and

L3 are similar in pattern to groups fed by Kartman (1953a), but this

could easily have been coincidence, and the prelarvae rank is exactly

the opposite of where it should be based on such a supposition. The

percentages positive for the 3 species (Table I- 2) are not significantly

higher or lower than those reported in the literature. Thus, 0.1% of

Cx. quinquefasciatus were found infected by Rosen (1954), although he

found none harboring L3; however, with Cx. annulirostris, he found that

2.5% harbored developing forms. Basio and Cheong (1977) reported rates

of infection varying from 0.03 to 0.82. for 4 different mosquitoes,

while an unusually high rate of 15.3% was reported for Mansonia

uniforms (Intermill and Frederick, 1970), although collections of

engorged mosquitoes from sentry dog compounds undoubtedly facilitated

improved rates. For Cx. quinquefasciatus, Villavaso and Steelman (1970)

reported a range of 0.0% (rural yard) to 2.jo (suburban), and there was

considerable variation within site types, depending on proximity of

breeding places and habits of dog owners regarding leaving animals


It is provocative that 776 (10/13) of all D. immitio infected

mosquitoes were found at site 1 (Table I- 3 ), which accounted for only

33% of all mosquitoes collected, compared with 51% for site 3, which

contributed only 15% (2/13) of the infected mosquitoes. While site 1

might be expected to have more infected Ae. taeniorhynchus because 58%

of this species was found there, and the rate at site 1 for APe.

taeniorhynchus was 7/155 (1/22) or 4.t, the total Ae. taeniorhynchus

at site 3 was 79, making the expected number of infected Ae.

taeniorhynchus at a rate of 4.6U at least 3 mosquitoes. However,

the only other site with any infected Ae. taeniorhynchus was site 2.

In a similar vein, site 3 accounted for 122/200 (61%) of all Cx.

quinquefasciatus; however, only one Cx. quinquefasciatus (3) was

found there, infected with prelarvae, while 2 infected Cx. quinquefasciatus

were found at site 1 with only 49/200 (24.5%) of all Cx. quinquefasciatus

collected. Thus, the abundance of the species was not related to its

infection rate, but seems, rather, to be site-dependent; site 1 is

clearly contributing a disproportionate amount of infected mosquitoes,

considering the numbers collected, or, conversely, site 3 is contributing

far fewer than would be expected if rate was a function of collection

size. To speculate for a moment, while site 1 was characterized in the

Materials and Methods in terms of its proximity to the Indian River

salt marsh, it also differs from the other 2 sites with respect to the

East Florida Ridge, a limestone spine running down the east coast of

Florida at varying distances from the Indian River or Atlantic Ocean.

It is also within 0.5 km of a golf course with pond water hazards and

much luxuriant, well-watered vegetation, thus providing both breeding

and resting areas for Cx. quinguefasciatus and dispersing Ae. taeniorhynchus.

Both other sites are west of that ridge, where the character of the flora

differs from that east of the ridge, possibly as a function of humidity.

Basio and Cheong (1977) saw that shady biotypes were more infectious.

While no data were collected in this regard, it offers habitat-oriented

alternative to the other salient possibility, mainly that there were more

infected dogs at site 1. Remembering that the weather conditions were

arid during most of the collection period, with rains coming toward

the end, the survival value of the more moist habitat at site 1 could

be extremely important, and offers an explanation for the observation

that 8C0 (4 out of 5) of mosquitoes with L3, likely, but not prima

facie, the oldest of the group, since it takes a minimum of 10 days to

develop L3, were collected at site 1. (Of course, a very old mosquito

can take its 1st bloodmeal and harbor only prelarvae on inspection).

That the dry spell produced an older population from which the samples

were drawn is fairly obvious. Not only was the population low, but,

without the influx of young mosquitoes, especially Cx. quinquefasciatus

and Ae. taeniorhynchus, which would not benefit from grove irrigation

practices as Cx. nigripalpus, the population would contain more and more

mosquitoes of older and older age and thereby increase the chances,

all other things being equal, of sampling an L5 infected mosquito, while

the portion of such mosquitoes would be diluted by large numbers of new

arrivals diluting lowering the average age. Survival rates in mosquito

populations are very low under good conditions (Nayar, 1980), and under

arid conditions, must be quite a lot lower; if the number of micro-

habitats conducive to survival were greater in one area than another,

it could be a major factor in disease transmission.

Assessing the vector potential of mosquito species naturally

collected but non-parasitized (equals negative results) is a common, if

risky, practice, based as it is on circumstantial inference of widely

varying strength, and no such assessment or judgment will be made on

this group here. However, since these inferential conclusions are

often cited in the literature, a discussion of their merit and application

to the present work is in order. The following parameters have been

used to support negative data (absence of infected mosquitoes from

wild collections) or to rank those mosquitoes found naturally infected: 1.

distribution (Rosen, 1954; Arnott and Edman, 1978); 2. abundance

(Rosen, 1954; Bemrick and Sandholm, 1966; Christensen and Andrews,

1976; Arnott and Edman, 1978; Lewandowski et al., 1980); 5. survival/

longevity (Rosen, 1954; Christensen and Andrews, 1976; Arnott and

Edman, 1978- mentioned in text but not included in table; Lewandowski

et al., 1980) 4. vertebrate host preference (Rosen, 1954; Bemrick and

Sandholm, 1966; Christensen and Andrews, 1976; Arnott and Sdman, 1978;

Lewandowski et al., 1980); 5. survival with heavy parasite infection

(Rosen, 1954; Bemrick and Sandholm, 1966; Christensen and Andrews,

1976); 6. multiple generations (Arnott and Edman, 1978); 7. too small

a sample (Christensen and Andrews, 1976; Bemrick and Sandholm, 1966);

8. seasonal abundance (Bemrick and Sandholm, 1966; Arnott and Edman,

1978; Lewandowski et al., 1980); 9. experimental vector potential

(Rosen, 1954; Beinrick and Sandholm, 1966; Arnott and Edman, 1978);

10. natural vector potential and infection rate (Rosen, 1954; Arnott

and Edman, 1978). In addition, mosquitoes have had their vector

potential reduced or eliminated for 1) insufficient numbers to be

reliable vectors, erratic populations, distribution is erratic,

not common, occurring sporadically later in the season, distribu-

tion.is pocketed, and late seasonal build-up (Arnott and Edman,

1978). The strength of the judgment could be enhanced when different

circumstantial lines of evidence are used in concert, as has often been

done, rather than eliminating a species because of the presence or

absence of one factor, unless that factor is ironclad. The problem

with this kind of assessment is that not all the parameters are

equally valid, and yet it would score the same on a yes-no type

ranking. For example, host preference is often cited, but this is

parameter of varying value, from weak to strong, depending on the

host and mosquito. The preference for bird blood shown by Culex spp

is often only just that, a preference, and many Culex readily feed on

mammals when mammals are abundant (Edman and Taylor,1968); thus, host

preference can be related to host abundance. On the other hand,

some mosquito host-preferences are more fastidious, such as with Cx.

territans, which will feed on frogs and other amphibia (Edman, 1974).

Most common mosquitoes, though, are blatant opportunists, including

Cx. nigripalpus (Edman, 1974), and the literature is replete with

reports of Culex feeding on dogs (see Villavaso and Steelman, 1970),

so the "weight" given to a parameter such as host preference is variable

and requires both knowledge and sense to insure its value to evalua-

tions of vector potential.

Likewise, it is not easy to determine what weight should be

assigned voltinism, or number of generations. A relatively large

number of generations is only important if it leads to an increase in

population, and this ir-rease in population is related to vector

potential. Since populations can go up or down with each succeeding

generation, depending on the survival and number of eggs laid by the

previous generation, the survival of which depended on environmental

conditions extant during their lifetime (Nayar et al., 1980),it is a

difficult parameter to correlate with vector potential. Abundance

itself may not be important if the vector is long-lived and is"ecological-

ly focused (Weinmann and Garcia, 1974), while, for normally rural

species, such as Cx. nigripalpus, abundance might be requisite for

vector potential, huge populations having to build up before sub-

stantial, epidemiologically-important penetration of residential

areas occurs.

Seasonal fluctuations in susceptible populations can be of

extreme importance in determining vector status. For example, in

the present investigation, only 27 Ae. vexans were captured, and

the very low rates of infection found in natural populations by

Bemrick and Sandholm (1966), 1 infected Ae. vexans per 859 collected,

mean that only extremely good fortune would have yielded an infective

mosquito from the sample if the same rate operates here. However, since

large Ae. vexans populations can rapidly be produced in groves during

periods of irrigation, if the survey had been carried on, it might

have included such a population and yielded infected individuals, if

incidence in this mosquito is related to abundance. Essentially, what

this implies is that sample-size can be very important, as noted by

Christensen and Andrews (1976), and, unless rates of infection are

extremely high due to focusing (Weinmann and Garcia, 1974), many

mosquitoes will have to be collected before a reliable and reasoned

judgment can be made. Thus, there is much more ground for Arnott and

Edman's (1978) claim regarding the low potential of Ae. vexans than if

the same claim were made for Ae. vexans here, because their sample size

(410) was considerably larger, although their "No" ranking of Ae.

vexans regarding field isolations ignores the work of Bemrick and

Sandholm (1966), and no "longevity" or "survival after infected" column

appears in their table of Biological Characteristics, although they

themselves acknowledge its importance in the text.

During the interval during which this survey was conducted (April

to August, 1977), it can be concluded that Ae. taeniorhynchus was the

major vector of canine heartworm disease in the Vero Beach, Florida,

area. Cx. quinquefasciatus was also an important vector and Cx.

nigripalpis,playing much the same role as in St. Louis Encephalitis

transmission, became entangled in the heartworm epidemiological web

by invading residential areas. Ae. taeniorhynchus has many of the

characteristics of Ae. trivittatus, which "carries very large numbers

of parasites, yet is able to retain the necessary mobility to fly". P. 278

(Christensen and Andrews, 1976). Furthermore, the large developing-

form loads in the tubules "suggests considerable tolerance by this

mosquito to the damage done by the developing juveniles." If "tolerance"

means survival, then the results are similar to those reported by Nayar

and Sauerman (1975) for Ae. taeniorhynchus, where survival of infected

mosquitoes for 15 days was 54%. The distribution patterns of CHD in

the eastern United States (Otto and Bauman, 1959) strongly suggests

salt marsh mosquito involvement. Ae. sollicitans, the predominant salt

marsh mosquito of the New Jersey and much of the northeastern coast, has

already been shown to be susceptible to heartworm disease (Beam, 1965),

although other studies (Nayar and Sauerman, 1975) indicate that its

vector potential is less than that of Ae. taeniorhynchus because of

poor survival (44) of the infected mosquito. Survival after infection

is of overriding importance to vector potential, and is one reason given

for the low vector potential of Ae. aegypti (Rosen, 1954). As to whether

Ae. taeniorhynchus would have been the dominant potential vector in a

succeeding collection period, or at any other point in time, is not

answerable by this study. Certainly, the seasonal fluctuations of mosquito


populations generates an array of relative incidences of infected

individuals of each species and possibly their rank in the order of

potential vectors.



Literature Review

That events occurring in the midgut of the arthropod host might

determine the portion of onchocercid miorofilariae that attain lodging

in the target tissue and ultimately become infective was first suggested

in experiments by Noe (1908), who observed the death of microfilariae

of Dipetalonema grassi in the gut of Rhipicephalus sanguineus, the

brown dog tick, during the final phases of feeding. Given the length

of time transpired since the initiation of feeding, it is not incon-

ceivable that digestive processes were involved. The digestion of

blood food by hematophagous insects is a complex subject, well-reviewed

by Cooding (1972), who states as a rationale for comparative studies,

"A second and more practical reason is that many blood-sucking arthropods

are vectors of pathogenic organisms. Since most of these pathogens

spend some time in the gut of the vector, digestive processes of the

insect could influence vectoring ability" p.5. Evidence in support of the

role of the midgut in mediating the intensity of filarial infections

in mosquitoes is suggested in the results of experiments by O'Connor

and Beatty (1938), who found only dead microfilariae of Wuchereria

bancrofti in a substantial number of Culex quinquefasciatus midgut

bloodmeals. Some mosquitoes harbored substantial infections but many

had "feeble" microfilariae. Similar results were described by Kotcher

(1941), who observed that a majority of Foleyella microfilariae were

killed in the gut of Cx. pipiens and Cx. quinquefasciatus. This work

was supported and expanded by Benach and Crans (1975), who reported

that only a small percentage of F. flexicauda penetrated the midgut

of Cx. territans, with most microfilariae dead in the midgut shortly

after feeding; the authors conclude, "Events occurring in the alimentary

canals . appeared to be the major cause for differences in vector

susceptibility" p.209; further, that "there are strong indications

that a gut barrier exists in Cx. territans which limits the number

of microfilariae entering the hemocoel. The lysis of the frog

erythrocytes and the death of the bulk of microfilariae, within a

short time after ingestion, suggest that this barrier may be related

to either the digestive secretion of the salivary glands and midgut,

or to mechanical action of the bucco-pharyngeal apparatus, or both"p211.

Although reports of microfilarial morbidity in the midgut are most

consistent in Culex, essentially the same pattern of events apparently

unfolds in the Anopheles ambiae:Brugia patei vector-parasite relation-

ship, very few microfilariae reaching the thoracic musculature, with

normal development of those that do (Laurence and Pester, 1961);

Crans (1973), comparing the susceptibility of An. gambiae and Cx.

quinquefasciatus, reported similar results, with far greater suscepti-

bility in the latter; differences in numbers ingested were not considered

a mediating factor in susceptibility as measured by intensity of in-

fection. In groups of Cx, quinquefasciatus fed on heparinized blood

infectious for B. pahangi, as much as 99.7%' of the microfilariae are

still in the midgut at +1 and, more importantly, this portion remains

essentially unchanged when measured at +24 (Ewert, 1965); moreover,

in terms of vitality, only one-fourth of the ingested microfilariae

showed any activity at all, and this was always feeble. Complete

refractoriness to infection by B. pahangi with dead microfilariae

in the gut has been reported in the same mosquito (Ogunda, 1969);

and, in the same system, using Cx. pipiens, Obiamiwe (1977) observed

that 99.1% of the filariae failed to escape the midgut; 2 barriers,

one relating to clotting time and the other to unknown factors, were

mentioned as sources of the observations. Working with the unsheathed

onchocercid, Dirofilaria repens, which does not penetrate the wall of

the midgut, but rather migrates directly to the Malpighian tubules,

Coluzzi and Trabucchi (1968) associated the death and injury of filariae

to the mechanical action of the bucco-pharyngeal armature, anatomically

known as the dorsal pharyngeal valve (Uchida, 1972). Dead larvae were

associated with mosquitoes possessing this structure, which includes

members of the subgenera Cellia and Nyssorhynchus of Anopheles and

Culex of genus Culex (only 1 Culex was examined, Cx. pipiens). These

experiments were expanded by Bryan et al. (1974) and McGreevy et al.

(1978), using B. pahangi and Wuchereria bancrofti with Cx.
quinquefasciatus, An. gambiae and An. farauti, and Ae. togoi and

Ae. aegypti.

Similar patterns of susceptibility indicating the presence of a

gut barrier have been detected in the D. immitis:Culex vector-parasite

relationships, indicating again that such barriers are not parasite-

specific. Evidence includes the observation by Feng (1930), that none

of 31 Cx. pipiens examined was susceptible to D. immitis, and the

comparative experiments of Hu (1931), the first researcher to discuss

mechanisms that might explain susceptibility variation amongst species;

he found "degenerate" microfilariae in Cx. ipien one day after

feeding, and concluded "Although it is a question what role physio-

logical and hereditary factors may play in determining the variation

in degree of resistance to infection with this parasite displayed by

the mosquito hosts, they must be of some significance" p.626. O'Connor and

Beatty (1938) found large numbers of dead D. immitis microfilariae in

the midgut-bloodmeals of Cx. quinquefasciatus bloodfed directly on the

infectious dog, while Cx. tarsalis, Cx. territans and Cx. pipiens

vector-parasite relationships with D. immitis were all placed in the

category of low intensity but supportive of normal development, in a

scheme of susceptibility levels devised by Yen (1958), including 5

classes and 7 groups, based on intensity and percentage of infection.

Yen concluded the question of why some species of mosquitoes are more

suitable as the intermediate hosts than others appears to be unexplain-

able by the morphological and taxonomical differences of mosquitoes,

for even members of the same species vary to a considerable degree in

their susceptibility to infection with the parasite. What mechanism

is involved in bringing about these variations in host susceptibility

of the different species of culicid hosts is difficult to determine

until we have a better understanding of the physiology and pathology

of the insect. Travis (1947) found dead D. immitis microfilariae in

Cx. quinquefasciatus, as Seeley and Bickley (1974) did in Cx. salinarius.

The earliest work in which the main focus of the study was to

uncover mechanisms to explain the variation in susceptibility level

amongst species was Kartman's (1953a) investigation of the vector-

parasite relation between D. immitis and 5 mosquitoes, Cx. quinquefasciatus,

Ae. aegy2ti and An. guadrimaculatue; salient events in the midgut were

specifically studied on a chronological basis, death of ingested

microfilariae being seen as early as 8h post-feed, leading Kartman

(1953a) to conclude, "The present work has indicated that the fate of

the microfilariae of D. immitis in the mosquito midgut is undoubtedly

a critical one for the completion of the parasite's life cycle. The

preponderance of negative females in the Culex species is explained by

the fact that the majority of the microfilariae are killed in the midguts

during the first 24h after the infective meal" p.69. Kartman (1953a)

ultimately concluded that the differences might be due to the presence

or absence of rapid bloodmeal coagulation, in turn tied to the secretion

of salivary gland anticoagulant, which would increase the exposure time

of Lmicrofilariae to antiblastic substances from the midgut or salivary

glands. In another study, the high resistance of Cx. nigripalpus to

D. immitis infection was attributed to the entrapment of microfilariae

in the rapidly-clotting bloodmeal and wounding of microfilariae,

possibly from needle-sharp crystals of oxyhemaglobin, formed as the

result of hemolysis of ingested red cells (Nayar and Sauerman, 1975).

As with Ox. quinquefasciatus in Kartman's (1953a) study, the low

intensity of infection for the group was attributable to the large

portion of refractory (negative) mosquitoes.

The gist of this body of research is that variation in susceptibility

to filarial infections in most Culex and some Anopheles is a function of

events first manifest in the midgut, which in most, but not all, mosquitoes

result in morbidity of most and sometimes all ingested microfilariae

shortly after the mosquito feeds. These studies suggest an array

of possible sources of the morbidity, including toxic factors, trauma

or entrapment, or combinations of these. Furthermore, this variation

occurs not only between species but within species; for example,

Hu (1931) found 1 Cx. territans with large numbers of prelarvae,

although a vast majority of individuals in the species support only

extremely low-level infections. Yen (1938), quoted above, alludes

to this same phenomenon, while Kartman (19538) states "the phenomenon

of individual response to infection thus becomes of particular signifi-

cance ." and talks of "especially susceptible individuals" p.66, and

Nayar and Sauerman (1975) report their results in terms of proportions

of the sample with a particular level of susceptibility. Thus,

variation in susceptibility levels of individuals is itself a character-

istio of the vector-parasite relationship in general, as is the normal

development of filariae once the gut barrier is successfully surmounted.

Although microbes, including bacteria and riokettsia-type or-

ganisms, have been observed in the bloodmeals of Cx. quinquefasciatus

(Chao and Wistreich, 1960; Micks et al., 1961), and that specific

functions associated with digestion of the bloodinal have been attributed

to microbes (Arnal, 1950; Ferguson and licks, 1961), with alterations

in the pattern of digestion resulting (Micks and Ferguson, 1961), no

evidence has yet been found which associates microbes with filarial

morbidity in any causal manner, although it is quite conceivable that

occasionally microbes could cause a reduction in susceptibility by

creating an unusually toxic environment in the midgut. However, the

midgut in sugar-fed mosquitoes is never distended as during bloodfeeding,

storage of sugar being the function of the diverticulae; only small

amounts of crop contents being passed to the midguts at any one time,

and, thus, the volume of liquid medium is never very large. Terzian

et al. (1952), and Terzian et al. (1955), had reported alteration of

normal susceptibility patterns in Ae. aegypti infected with Plasmodium

gallinaceum, but they specifically attributed the effects to disturb-

ances of the host physiochemical homeostasis, which produces the normal

pattern. Thus, aside from what seem to be alterations of the suscepti-

bility level of mosquitoes to Plasmodium induced by midgut environmental

changes, with perhaps additional potential mechanisms from the absence

of competition for nutrients, or absence of factors produced by the

bacteria and used for host defenses (icks and Ferguson, 1961), no

antiblastic attribute has been assigned bacteria even in the case of

Plasmodium, and it is unlikely that digestive enzymes, from either

host or flora, which reach peak activity (cooding, 1972) too late to

be involved in the initial destruction of filarial parasites, are

involved. Thus, the role of bacteria in mosquitoes, even as symbionts,

is most likely associated with the crop in sugar-fed only mosquitoes,

while with regard to the midgut bloodmeal, exert their influence only

when enough time has passed to allow the population to build up. In

fact, more than a third of An. crucians, 6Q% of An. quadrimaculatus and

10Q/ of Ae. infirmatus examined for microbes were sterile (Hamilton,

1975), who aleo reported an absence of association of bacteria and early

development of D. immitis in monocontaminated An. quadrimaculatus.

Normal, albeit sluggish, microfilariae of D. immitia are seen in

+24h bloodmeals in Ae. aegypti, and it is questionable whether intact

microfilariae are subject to enzymatic attack in mosquito midguts at

all (Kartman, 1953a). Thus, the inherent inconsistency of microbial

infection of mosquito midguts, compared with the consistency of the

observed morbidity, does not support the contention that flora are

involved in any significant way, even as sources of hemolysis (Ferguson

and Micks, 1961) or coagulins, in the determination of susceptibility

to filarial infection.

The vector-parasite relationship between Ae. aegypti and

D. immitis is consistent only in the variability of the elements used

in its characterization; this is a property of the Ae. aeypti:filariid

relationships in general. The first of these elements is the decrease

in survivorship of mosquitoes that ingest large numbers of micro-

filariae, first observed by Fulleborn (1908). Following this study,

many reports using Ae. aegpti in conjunction with a variety of

onchocercids, including Brugia (Wharton, 1957), Dirofilaria tenuis

(Pistey, 1958) and Wuchereria (Rosen, 1954) have alluded to this phenomenon,

and specific studies have even focused on it (Kershaw et al., 1953; Weiner

and Bradley, 1970). Travis (1947) reported extremely high mortality rates

for Ae. guamensis and Ae. pandani after ingestion of D. immitis; nearly

identical results were reported by Rosen (1954) for Ae. edgari and Ae.

polyneneis infected with D. immitis on Tahiti. These studies support the

view that failure to survive after ingesting even moderate levels of micro-

filariae, while highly variable in expression, is actually one of the

more consistent aspects of this vector-parasite relationship and serves

as an excellent example of the reversibility of the epidemiological

triad; not only does the vector affect parasite but the parasite can

affect the vector, and this, in turn, can very significantly alter the

vector potential of an infected population. Vector mortality is not

specific to this mosquito-it occurs in several other mosquitoes in

other genera (Weiner and Bradley, 1970; Hamilton and Bradley, 1979)-

and it does not occur in all Aedes; for example, it is not a property

of the vector-parasite relationship between D. immitis and Ae. triseriatus

(Intermill, 1975) or Ae. taeniorhynchus (Nayar and Sauerman, 1975).

The most frequent observation is that the midgut epithelium is being

destroyed by the writhing and lashing of microfilariae, with rupturing

of the cells and contamination of the hemocoel with gut contents,

mortality from septicemia or toxemia ensuing.

The second element used to characterize this relationship, first

reported by Roubaud (1937), is the variation in susceptibility level

to D. immitis infection between geographical strains of Ae. aegypti,

as confirmed by Kartman (1955a). Strain variation also characterizes

the relationship between Ae. aoegpti and other onchocercids, such as

B. malayi (MacDonald, 1962), and Wuchereria, Brugia and Dirofilaria

(MacDonald and Ramachandran, 1965; Beckett and MacDonald, 1971;

Obiamiwe, 1977). A single, sex-linked, recessive gene, designated

I"f"", was found to control susceptibility level to Brugia and Wuchereria

but not Dirofilaria (MacDonald, 1962; MacDonald and Ramachandran, 1965).

Failure of penetration of the tubule cells in Ae. aegypti by D. immitis

microfilariae was an expression of non-susceptibility in one genetic

study, which concluded that other factors, such an encapsulation,

melanization and defecation, were not mechanisms mediating refractori-

ness (McGreevy, 1972).

The third element that can be used to characterize the host-

parasite relationship between Ae. aegypti and D. immitis is the arrest

of development of prelarvae after attainment of lodging in the target

tissue, the Malpighian tubules, first described by Roubaud (1937),

although Hu (1951) alludes to this phenomenon briefly. Yen (1938)

reported "retardation of growth" and presence of a prelarva in the

Malpighian tubules on the tenth day post-infection in Ae. stimulans

and Ae. canadensis, but normal development in most Ae. cinercus and

Ae. vexans; it is of interest that this researcher observed the

phenomenon in Culiseta inornata and Coquillettidia perturbans. It

is not characteristic in Anopheles. Travis (1947) notes that in

Ae. aegypti large numbers of larvae migrated to the Malpighian tubules

from the stomach, and that in this species there was a much larger

drop in the numbers that reached the sausage stage than in the other

two species of Aedes, while Kartiran (1953s, p.3j) remarks "the Ae. aegypti

females showed a predominance of unchanged micrcfilariae in their

Malpighian tubules and this phenomenon was a daily finding throughout

the experiment," which contrasted with his findings on the other

mosquitoes investigated (An. quadrimaculatus and Cx. pipiens and

Cx. quinquefasciatus); further, "Some of the unchanged microfilariae

in aegypti were still alive fifteen days after being ingested'! p,33,

observations corroborated by Nayar and Sauerman (1975).

The cause of this arrest has been speculated upon and there are

several hypothesized mechanisms, each supported by their own evidence.

While one of these, encapsulation, is often considered a separate

element of the host-parasite relationship, probably because the

phenomenon is so widely observed in mosquito-filariae relationships,

there is no evidence, other than studies using light microscopy

observations, that definitely eliminates it from involvement. Pigmental

encapsulation of D. immitis occurred in Ae. vexans Malpighian tubules

studied by Grassi and Noe (1900), and in all Aedes studied by Yen (1958),

and he considered this to be the source of developmental arrest in Cu.

inornata and Cog. perturbans. Yen (1938) concluded that the capsules were

the "shed skin" of the parasite "the products of degeneration," and that

all capsules are within cells, and he disagreed with earlier reports

alluding to a chitinous chemical nature in the capsules. Kartman

(1953a) notes that "the invasion of Ae. aegypti by D. irmitis was

attended by a consistent encapnulation of the microfilariae, . ." p.35,

but only 12% of the mosquitoes dissected harbored these forms, and he

concluded that the phenomenon was ". . of little quantitative

significance" p.35. However, Lindemann (1977) describes 1090

encapsulation of microfilariae of D. immitis in the tubules of Ae.

aegypti, although the infectious load was relatively light.

Another possible mechanism is the laying down of membranes by

the host around the body of the parasite. Membranes can play a very

important role in host-parasite relationships involving intracellular

parasites, with membranes of both host and parasite origin being laid

down (Traeger, 1974); Ashley (1972) observed such membranes around

D. immitis in Ae. aegypti tubules using electron microscopy, while

more recent comparative studies along similar lines indicate that the

phenomenon might be associated with developmental arrest (Bradley and

Nayar, 1980). Since the parasite is predominantly intracellular in the

early stages (Taylor, 1960), relying on trans-cuticular absorbtion of

nutrients (Nelson, 1964), the membranal isolation of the parasite offers

an attractive hypothesis to explore. However, if the membranes are

lysosomal, the failure of prelarvae to be digested and the presence of

living but moribund prelarvae up to 15 days post-feed must be ex-

plained. Furthermore, as with encapsulation, the cause-effect aspect

must be resolved, since enveloping of forms made moribund or killed

by other mechanisms is quite possible, based on the current knowledge.

Aside from physical isolation mechanisms, possible chemical

mediation of arrest of development is indicated from the work of

Weathersby and McCall (1968) and Weathersby et al., (1971) who found

that substances in whole body extracts could alter susceptibility

levels in Ae. aegypti to Plasmodium gallinaceum infections, factors that

were apparently "antiblastic" in action. Atreptic (nutritional) sources

of arrest are suggested in the work of Traeger (1974), who demonstrated

a requirement by Plasmodium lophurae for adenosine triphosphate, but

not as a purine, since the diphosphate form was not efficacious, but,

more likely, as a co-factor in membrane transport. In the absence of

the correct chemical substances, the prelarvae could be placed in a

stress situation, continuing their lashing, thereby destroying the

structure of the target tissue.

Finally, the absence of requisite physical-chemical conditions

in the tubules of Ae. aegypti might be the actual mechanism that causes

the arrest; if the parasite requires a quiescent period prior to

initiation of development, the disruption of the composition of the

tubule cells by the filariae due to an inherent structural factor in

these cells such as relatively fragile membranes, or the absence of

proper turgor osmoticc pressure), could explain ultimate arrest. It

is understandable that such microfilariae would be under a continual

stress, unable to pass into the quiescent state. External physical

factors could affect development of the parasite, including temperature

and pressure; humidity has been associated with varying levels of

susceptibility (Basu and Rao, 1939).

The work of Roubaud (1937) suggests that the phenomenon of arrest

of development is not universally distributed amongst the varying

strains of Ae. aegypti, with susceptibility ranging from 100 percent

to nearly total refractoriness, and it was logically the arrest of

development that was involved in the genetic expression of decreased

susceptibility, as suggested in the work of Kartman (1953a) and

McGreevy (1972); McGreevy et al. (1974).

This historical review suggests that there is at least a general

association between the type of mechanism that mediates susceptibility

to filarial infections and the genus of the vector. The fate of the

parasite can be decided at the gut level or target tissue level, as

suggested by Nayar and Sauerman (1975), but apparently not both in the

same mosquito. aFrthermore, the expression of these mechanisms is

variable within species and, again, within strains. Regardless of

site of action, the net result of both mechanisms is the reduction of

host efficiency and, ultimately, vector potential.

Definitions and Usa9ae

In general, the terminology of Kartman (1953a) is employed here,

with the following modifications and additions:

icrofilariae (MF)- generally considered precocious embryos

because they lack the functional alimentary tract characteristic of

other nematode first-stage larvae (Taylor, 1960), in the present study,

this is the form that occurs in association with vertebrate blood, which

includes, therefore, both the vertebrate bloodstream and invertebrate

gut. This form has no potential for development in either the vertebrate

host blood or invertebrate alimentary canal.

Prelarvae (PL)-- the precocious embryoes, indistinguishable from

microfilariae, but not found in association with the blood, having

attained lodging in the target tissue, in the case of D. immitis, the

Malpighian tubules. This form has the potential for development on the

basis of environment. The terminology for microfilariae and prelarvae

in the literature is inconsistent and confusing. To give just a few

examples, Kartman (1953a) uses the term microfilarial to refer to all

pre-developing stages, as does Taylor (1960), while Pistey (1958)

assigns the term prelarvae once they have been ingested by the arthropod,

which is closer to the usage here. Other criteria encountered were

"G"-cell division, shortening, thickening; thus, the issue seems more

one of semantics than biology. The present terminology is both logical

and functional; logical, in that microfilariae have no potential to

develop into larvae (no development of microfilariids while in associa-

tion with the vertebrate blood has been reported) while prelarvae,

having attained the invertebrate target tissue, do have that potential;

a larva is a developing stage, with the prelarva its precursor;

functional, in that each term is restrictive as to location, prelarvae

never occurring in the gut.

Developing Form (DF)-usually, as in Kartman (1953a), any stage

beyond prelarvae; in the present study, in the case of Culex nigripalpus,

the prelarvae is equivalent to the developing form, since no barriers

beyond this stage have been demonstrated in Culex:filariid relation-

ships; undeveloped pre]arvae are not normally found at +120h in moderate

infections. In Aedes aegti, this relationship does not hold; that is,

prelarvae and developing forms are not equivalent, because undeveloped

but living prelarvae are often found in the Malpighian tubules at +120h

and even +240 .

Suceptible (Mosquito) --a mosquito that supports the development of

any filarial larvae (even one). This definition is far more inclusive

than that of Hu (1931), where an individual was considered positive for

susceptibility only when all the larvae seemed likely to complete

development or already had; this usage would eliminate those mosquitoes

harboring developing forms concurrently with undeveloped prelarvae;

since these mosquitoes can become infective due to the former group,

it is not logical to assign them a category that would imply that they

are not susceptible. In the present case, the word "susceptible" is

inclusive and does not designate a specific level of susceptibility, as

"refractory" (see below) does. In this study, the word "susceptible"

is used with or implies the presence of a qualifier for designation of

a particular level,such as, "low-level," "moderate," "Anopheles

guadrimaculatus level," of susceptibility, and so on; thus, it is a

relative term, covering an array of situations, the only thread

connecting them being the presence of at least one developing form,

regardless of the number ingested.

Refractory (Mosquito) -a mosquito harboring no developing forms,

regardless of number of microfilariae infested.

Atreotic/Antiblastic (Weathereby and McCall, 1968)--terms used to

designate the source of factors causing an event, the first referring to

nutritional factors that operate by their absence; that is, the absence

of an essential factor, such as an amino acid or vitamin; the second

refers to factors that operate by their presence, such as general and

specific toxins.

Materials and Methods

Entomological and Parasitological Materials and General Procedures

Biological material

The mixed breed, 18 kg male dog used in this study to both feed

mosquitoes directly and to provide a source of microfilariae for other

studies harbored a naturally-contracted heartworm load that produced a

microfilaremia of 75-150 microfilariae/ul when sampled by venipuncture

of the cephalic vein at approximately 10 + 1hr White leghorn chicks

and chickens, aged from 2 weeks post-hatch to adults, were continuously

available at the Florida Medical Entomology Laboratory (FEEL).

Three mosquitoes, Cx. nigripalpus, Ae. aegypti and An. quadrimaculatus,

each available as colony material courtesy of Dr. J. K. Nayar of the

Florida Medical Entomology Laboratory, Vero Beach, Florida, where these

studies were conducted, were chosen, based on previous experimental

use (Nayar and Sauerman, 1975), to serve as models for particular

susceptibility patterns, associated according to the literature with

particular genera or portions of genera. The Cx. nigripalpus colony,

established by J. S. Haeger during 1975, possessed all the characteristics

reported for the genus, and was, furthermore, readily available year

round in case of colony failure. The Vero Beach strain of Ae. aegpti,

developed by J. K. Nayar during 1974, likewise produced arrest of develop-

ment and exhibited variable susceptibility levels and vector mortality.

The An. guadrimaculatus colony, origia:.l y from Gainesville, Florida,

was essentially 100 percent comprised of susceptible individuals, and

was, therefore, nonpareil as a microfilarial and prelarval vitality

control. Rearing procedures for each species were conventional and are

described elsewhere (Nayar and Sauerman, 1975). Cx. nigripalpus

used were between 7-21 days post-emergence, while An. quadrimaoulatus

and Ae. aegypti were between 7-14 days post-emergence, and all species

were maintained in 15 cm3 holding cages on 1C0 sucrose solution at

approximately 80( relative humidity, until 24 prior to a scheduled

feeding, when the sucrose was replaced with water.

Feeding of mosquitoes

For direct feeding on the dog, either the paw and lower leg were

placed in the cage through the stockinet sock, or the dog's chest was

laid down on top of the upward-facing screened side of the holding cage,

while its head and rears were supported by the closed upper sides of 2

other holding cages, padding added to make the animal as comfortable

as possible to reduce struggling (Figure 11-1,2). In general the

slowest-feeding of the 3 species, most Cx. nigripalpus fed within 30

minutes; they are far more tenacious once they have initiated probing

and are less likely to be dislodged by momentary movements or leg

twitches by the dog than the other 2 species, which, however, take their

fill more rapidly. Furthermore, engorging by Cx. nigripalpus is delayed

by the "pool-feeder" behavior, characteristic of many mosquitoes in the

Culex genus, which leaves a small red dot approximately a mm in diameter,

which take several days to disappear, as evidence of its visit; no such

marks are produced by the feeding of the other 2 species. After re-

moving and discarding unfed and partially-fed individuals, engorged

specimens were removed to a 270C bicroom where they were held until the

particular experiment had them scheduled for dissection.

For feeding mosquitoes via membrane, an apparatus was constructed

which incorporated the inprovements of Vade (1976) on the Rutledge et al.

Figure II-1. Feeding methods: direct feeding
legs of dog inserted through access sock.

of mosquitoes on

Figure 11-2. Feeding methods: direct feeding of mosquitoes on
shoulder of dog laid down on top of cage.


(1964), design with a direct-drive, variac-controlled stirring device, to

prevent sedimentation, and temperature control of + 0.1000 (Figure II-3

and 4); as little as 0.25 ml of blood could be used in the feeding

tube, which was fashioned from a length of thick, boiler, glass

tubing, and the entire waterbath-feeder tube-stirrer unit was sealed

at the top after charging the feeder, which kept the relative humidity

within the chamber near 100%, reducing evaporation of blood from the

tube. The entire feeding assembly was connected with tubing to a

5-gallon tank containing a thermostatically controlled, 120-volt

nichrome wire water heater and submersible, variac-controlled, "Little
Giant"R pump, along with the thermo-regulator and thermometer. Cx.

nigripalpus fed best on media warmed to approximately 340C, while Ae.

aegypti and An. quadrimaculatus preferred 370C.

An array of membranes was subjected to preliminary testing for

suitability for use as the feeder tube end-cover, including hog-gut

sausage casings, Baudruche, chick skin, chicken or turkey crops,

Parafilm, prophylactic contraceptive rubber, and Fourex natural

animal membrane prophylactic contraceptives. While As. aegypti fed

best on cleaned chick crops, which was also Rutledgeet al.(1964) best

membrane, Cx. nigripalpus definitely preferred the FourexT (washed

thoroughly before use to remove factory lubricants/preservatives),

and this was fortunate, as each contraceptive yielded 8-10 high quality

(stripped of all fascia and fiber) and durable membrane covers for the

feeder, at approximately $0.10 per cover and far less time put into

preparation than chick crops. An. quadrimaculatus were more catholic,

although they showed a slight preference (preferences were not measured

experimentally, but differences were quite obvious) for the higher






D |


Figure 11-4. Feeding methods: membrane feeder in position on top
of cage, with water bath reservoir (left) containing pump and

feeding temperatures and chick crop membranes. Membranes were attached

to the feeder tube with a rubber grommet, the outside diameter of which

allowed a water tight fit to the feeder tube water bath, and 1-2 mm

of tissue was allowed past the end of the tube, creating a bulging

membrane just past the membrane-grommet interface. The adjustable

stirrer was set about a mm above the membrane bottom, past the end of

the feeder tube by a few mm, into the bulge caused by the weight of


For blood feeding mosquitoes, 0.5 or 1 ml of blood (the amount

depended on the number of mosquitoes to be fed) was drawn from a

suitable leg vein on the dog (all 4 legs were used when schedules

called for frequent feedings, but this did not amount to more than an

ml per day under daily feeding schedules) and placed in a 10 x 75 mm

test tube containing 0.1 ml saline (0.90% NaC1, unless otherwise stated),

for each ml blood and anticoagulant, unless glass bead-defibrinated

blood was used, in which case the 0.1 ml of saline only was used.

Three antiocagulants were used: 1-2 mg/ml blood sodium heparin, with

an activity of 155 units/mg (Nutritional Biochemicals Co.) (the minimum

amount used is more than 10X the amount required to prevent the blood

from clotting); 3.8 mg sodium citrate (0.389)/ml blood; 2 mg EDTA

(ethylenediamine tetraacetic acid)/ml blood. For glass bead-defibrinating,

the freshly drawn blood was placed in a 25 ml flask half-filled with 3 mm

glass beads (Sigma Chemical Co.) and with 0.1 ml saline; the stoppered

flask was attached to a rotator and the preparation churned for 30

minutes at 40-50 revolutions/minute; anticoagulant-treated blood was

either rotated or shaken every few minutes after more vigorous initial

shaking to insure distribution of anticoagulant; additives were mixed

with the blood simultaneously with anticoagulant. Chicken blood

drawn from the wing vein and jird blood from cardiac puncture were

processed similarly. Heparinized plasma was obtained using conventional

and centrifugation techniques; serum was obtained by placing a 5 ml

aliquot of freshly drawn blood into a Sigma Thrombin Vial (stock

number 850-1), mixing and incubating the vial for 2h at 370C. This

technique is more likely to convert all fibrinogen present to fibrin

than conventional techniques, which can allow fibrinogen to be present

after clotting since the plasma thrombin levels are variable. Washed

red cells were prepared by 3X washing in saline after initial centrifuga-

tion of anticoagulant-treated whole blood and decanting of plasma.

The fluids were added to the feeder tube via Pasteur pipette. The

feeder tube with membrane in place is maintained until this moment with

a few ml of saline in the tube and the membrane suspended by the grommet

in a saline-filled small dish; the saline is discarded and the tube

and inner membrane surface are dried with a length of rolled paper

towel, dabbing the membrane gently. After securing the feeder tube in

the bath and setting the stirrer length, the tube is brought to tempera-

ture via the water bath, the fluid added, and the apparatus placed on

top of the cage of mosquitoes, the membrane-covered tip of the feeder

tube barely inserted into the sugar vial hole. The stirrer is quickly

placed back down into the feeder tube,and the entire system sealed with

appropriately-sized plastic lids. The stirrer speed can be controlled,

but 100-200 revolutions per minute was the usual setting. No effect on

microfilariae from the paddle action was observed,and the results of

experiments are not indicative of vitality loss from this source.

Separation of microfilariae

Microfilariae free from the blood were needed for several pro-

cedures, and these were obtained by placing an aliquot of freshly

drawn anticoagulant-treated blood into a 15 ml centrifuge tube,

adding citrated (0.5%) saline up to 15 ml, mixing thoroughly, adding

10% saponin drop by drop 3-4 drops usually sufficing for less than

an ml of blood and checking for the typical clearing of the mixture

after the addition and mixing of each drop. As soon as clearing is

noted, the mixture is immediately centrifuged at 900 rpm for 5-6

minutes, the clear red fluid decanted and replaced with fresh, non-

citrated normal saline, the pellet agitated into the saline, and the

procedure repeated once more. The pellet of microfilariae is very soft

due to the relatively slow separation speed, and decanting was achieved

using a rubber bulb-type automatic pipetter on a 25 ml volumetric

pipette. If these microfilariac were to be suspended in another fluid,

the fluid was simply added to the final pellet and gently vortexed;

otherwise, if the microfilariae were to be used in saline, a few drops

of the saline was added into the centrifuge tube, vortexed gently and,

prior to use, transferred as a 100-200 pl drop to a siliconized slide,

then covered with a high humidity cap. Microfilariae in such drops

tend to aggregate in the center of the drop,facilitating collecting in

a concentrated form with minimal dilution from the saline, an important

consideration in certain procedures, such as testing of extracts. When

a procedure necessitated that microfilariae remain in non-blood fraction

fluids such as saline for more than Ih, the final washing and suspension

was done using Hank's or Locke's buffered salines. High humidity caps,

to curtail dessication of droplets, were fashioned from 3 ml-syringe

sterile container end caps which were atomized with saline solution

just prior to use; condensation droplets usually formed on the dry

area of the slide surrounding the drop under the cap within 50 seconds,

and continued vitality after several hours under such conditions

suggestshigh efficacy in the system, and droplets as small as 20 pl

could be maintained thusly.

Dissection for midRut microfilariae and tubule prelarvae

Conventional technique was used to remove insect midguts and

Malpighian tubules, the salient aspects of which will be described.

The materials of choice are 2 pairs of insect forceps, insect or

normal saline, siliconized microscope slides with 24 x 50 mm cover-

slips, and a 10-20X dissecting scope with accessories. With the

specimen ventral side up and the tip of the abdomen in a drop of

saline, the thorax is grasped with one forceps and subterminal segment

with the other, which is used to pull the hindgut and rectum into the

droplet. If the head is not removed, the gut usually disconnects at

the cardiac sphincter yielding the midgut, tubules and hindgut, while

sometimes only the hindgut and tubules are obtained when the breakage

occurs at the pyloris. With the head removed, with practice and if

care is taken, the 3 diverticulae are also obtained, the breakage

usually occurring at the esophagus, a short tube connecting the posterior

of the pharyngeal pump with the proventriculus. If the midgut is grasped

at the pyloris, and terminal or subterminal segments grasped and pulled

away, the disconnection usually yields the midgut and tubules, which are

easily separated by grasping the anterior of the midgut and pulling

the tubules away, grasping them as proximal to the gut as possible.

Although tubules can be examined individually on slides, each set with

its own coveralip, as many as 15 sets could be processed on the sane

slide by simply placing each set in its own 10 )l droplet (the initial

droplets can be far larger to avoid problems caused by dessioation;

but using lung-powered drawn capillaries in mouth aspirators, the

final volume just prior to coverslipping is quickly reduced to ap-

proximately 10 ul) and a 24 x 50 or 24 x 60 mm coverelip evenly

lowered onto the droplets. It is important not to have too much

liquid under the slip, since compression squeezes out contents from

under the slip and material can be lost,or preparations mixed,when the

pressure is released. In the non-compressed tubule, prelarvae are

usually quite occult, but compression and release disintegrates the

tubules, releasing the prelarvae into the medium in the vicinity of

the tubule debris. Unless so stated, all prelarvae in each tubule

were searched for and counted. Unlike sausages and later stages, the

prelarva is a durable creature and usually quite vital following

compression, thrasing and undulating in normal fashion (Figures 1-5,6).

Host efficiency assay

To determine the number of microfilariae ingested, the epithelium

is gently removed from around the bloodmeal and shaken to dislodge

parts of meals; this is a simple matter in Anopheles and Culex where

the meal is rapidly isolated by the peritrophic membrane, more difficult

with Ae. aegypti, in which no membrane forms for many hours, and

portions of the bloodmeal are embedded in crypts in the epithelium.

However, unless steps are taken to prevent it, the bloodmeal coagulates

rapidly in Cx. nigripalpus (Nayar and Sauerman, 1975), trapping many

microfilariae in the rubbery clot, in which visualization of the micro-

filariae is impeded. Again, as with prelarvae, the clot must be subjected


Figure 11-5. Malpighian tubule
in vitro bright field; x40.

squash for prelarval count;

Figure 11-6. Malpighian tubule squash for sausages; in vitro;
Nomarski optical systems; x100.

to compression to reveal the presence of the larvae or to release the

microfilariae into the medium, a compression of far greater force than

used for tubules, and even with this, the genuine clot is not dissembled,

but merely springs back into shape; it is this characteristic which

serves as proof of coagulation occurrence, since agglutinated meals,

after subjected to compression, do not re-coalesce but remain dissipated.

Interference with counting due to coagulation is not a problem in the

other species, if dissections are performed soon after feeding, and,

even if they are not, the number of trapped microfilariae is far less

than in Culex, where the problem can be eliminated only by feeding with

glass bead-defibrinated blood or microfilariae suspended in serum.

Interference with counting from red cells is eliminated by the addition

of a minute amount of (0.1 0.5y1 ) 10Q saponin into the droplet of

saline containing the meal, swirling with a lung-powered air jet through

a drawn capillary until the characteristic transparent red color of

hemolyzed red cells is seen. Using the oame capillary used for swirling,

the preparation is transferred to a modified Biles (1975) hemacytometer

type counting chamber, which was found after comparative testing to

out-perform the next best vital counting device, the Sedgewiok-Rafter

chamber, by a wide margin, and this device was used in connection with

a variety of procedures in this study. The only modification from

Biles (1973) was the use of No. 2 coverslip pieces as supports for

examining very small quantities of fluid (less than 1 il). The compres-

sion produced a flattened disc at least a few mm in diameter. For

larger volumes, the design was essentially that of Biles (1973). Not

only are there no obscuring edges in such a chamber, there is no tendency

for the liquid to travel (chambers must be siliconized to work efficiently),

and evaporation is minimal. One bloodmeal in saline could be separated

out into 100 2 pl droplets in the chamber and scanned efficiently and

quickly. Some losses due to adherence of microfilariae to the sides

of capillaries, which happens even when they are siliconized, undoubted-

ly occurs, but this is not significant when the microfilaremia is as

high as it was in this study.

The foregoing procedures generate several data types, including

number of microfilariae ingested per mosquito and average number of

microfilariae per mosquito per experimental group, number of prelarvae

per mosquito and average number of prelarvae per mosquito in a group,

proportion of prelarvae of total ingested microfilariae, called the

"host efficiency" index by its inventor, Kartman (1953a), and the

frequency distribution of prelarvae per mosquito in a group, which

can be used to form a "susceptibility" profile based on a comparison

with the susceptible model, An. quadrimaculatus. The advantage of the

prelarvae count is that it accurately measures the resultant of forces

manifested in the midgut, while avoiding difficulties and added variables

encountered with older infections. It is a characteristic of the Culex

genus in its relationship with D. immitis that once the gut barriers

are successfully surmounted, and prelarvae are safely in the tubules,

development proceeds from that point without significant impediment;

thus, with Culex, the prelarval count is a good estimator of vector

potential, since the developing form rates, used here to demonstrate

barriers past the midgut, was essentially the same as in the susceptible

model (the test does not discriminate amongst sources of poor host

efficiency, this being the purpose of the vitality assay, to be

described in a later section). Thus, it is this assay that produces

the frequency distribution of susceptibilities by which the species

is characterized. The most frequently occurring values of prelarvae

per mosquito for Cx. quinquefasciatus (Kartman 1955a) and Cx. nigripalpus

(Nayar and Sauerzan, 1975) are low to negative, and thus the mosquito

as a species is placed in a low susceptibility category regarding its

veotor-parasite relationship with D. immitis, while the opposite is

true for An. uuadrimaculatus. Further barriers beyond the gut are

characteristic of the Ao. aegypti:D. immitis vector-parasite relation-

ship, and, therefore, this assay is inappropriate for measuring

susceptibility in that species. With Ae. aegypti, alterations in

susceptibility level are detected using the developing form assay, in

which Malpighian tubules are removed and inspected for the presence

and portion of sausage forms at 120 + 24h after feeding. As in the

prelarvae per female assay for Culox, sets of tubules are subjected to

coverslip pressure; if considerable pressure is required to reveal all

forms, this cannot be discontinued until the count is complete, as

upon release, the soft-bodied sausages are no longer recognizable.

The allowable interval for dissection was from 96h to 144 after

feeding, but while later times could probably be used without loss of

information, efforts were always directed, in the present study, towards

conducting dissections at +120 Under normal experimental conditions,

most D. immitie will have "shortened" to sausages by +72h (Kartman,

1955a; Nayar and Sauerman, 1975), and preliminary studies indicated,

as McGreevy (1972) reported, that most prelarvae that fail to transform

to sausages by 48-72h are either moribund or dead. There is never any

sign of bacterial activity in the tubules, which are presumed to be

sterile, since dead prelarvae, while sometimes encapsulated, are not

digested by microbial activity.

Microfilarial (and prelarval) vitality

Normal microfilariae (includes prelarvae for purposes of this

discussion) appear as a vigorously thrashing and rhythmically un-

dulating organisms, the serpentine movements occurring concurrently

with the twisting and turning back on itself. This is also character-

istic of prelarvae up to about 48h after attaining the tubule environ-

ment. Two types of morphological appearance for morbid microfilariae--

dead and intact and dead and disintegrating-were described by Kartman

(1953a) as part of an assay. In this study, all immobile forms are

aggregated with those exhibiting bizarre movements into a single

category of "moribund," or dying, microfilariae; not included in this

category are sluggishly-moving but rhythmically undulating individuals,

a situation that most frequently occurs after they have remained in

the particular medium for at least 30 minutes and is evidence of a

non-supportive, but not necessarily toxic, environment where morbidity

is a function of the absence of required (atreptic) factors in proper

balance. On the other hand, if morbidity was induced within 10 minutes,

this was taken as a sign of the presence of antiblastic toxins, even if

normal but rapidly damping down undulations accompanied the loss of

vigor. Any interference at all with rhythmic undulations was interpreted

as the effect of physical or chemical trauma, since preliminary observa-

tions indicated that such larvae do not recover. Although there is

good agreement in the literature that total vacuolization of the micro-

filariae is synonomous with death, such larvae were still placed in the

more inclusive moribund category. The assay generates several data

types, including number of microfilariae or prelarvae that are morbid

in an individual mosquito, average number of moribund microfilariae in

a mosquito in a group of mosquitoes, and a vitality profile, since

vitality is not an all-or-nothing occurrence in the individual but

instead is a matter of proportions of moribund microfilariae in an

individual mosquito, which generates a frequency distribution for a

particular group, allowing a comparison of distributions without

losing sight of individuals. This is an all-important consideration

in the present study, since it is often the small portion of individuals

exhibiting behavior uncharacteristic in terms of the rest of the group,

to a point where it is clear that they represent a different treatment,

which are most illuminating in investigating the problem. The Biles

(1973) hemacytometer was used for all determinations, since this method

reduces factors that could affect vitality duo to atreptic deterioration

of the microenvironment. The only exception was direct ooverslip

compression preparation of extracts, to be described below, in which

no transfers were made.

Coagulation time

Conventional techniques were utilized to measure coagulation time;

a fibrin rake with tines approximately 1 mm apart was fashioned using

minute pins epoxied to a wood applicator stick. Detection of first

threads, which event serves as the basis for some automatic fibrinometer-

type devices, were easily detected, and the growth of the clot could be

monitored, which was necessary, since in cases where the amount anti-

coagulant is small, clotting can cease, making questionable its equati-

bility with fully clotted material. A high level of accuracy was not

necessary for this study and accuracy decreases with the number of

samples being simultaneously tested, if the number of operators remains

constant. For rapidly clotting preparations, fewer samples were monitored,

if accuracy was desired. The maximum number sampled at any one time

in this study was 6, 3 of which were controls, since controls were run

on the same slide as the experimental group in the event that impurities

on the glassware influence results. Each sample on a siliconized slide

was covered after raking, which consumed about 3 seconds, with a high

humidity cap, constructed as described previously in this section.


Chicken erythrocytes are nucleated,and when hemolysis of the cell

occurs, as with distilled water or saponin, the free nuclei liberated

into the medium serve as intact markers of the event. On Giemsa-

stained thin smears, solitary nuclei,which disintegrate as they dry

on the elide into an amorphorous round mass often with feathered edges,

stain the expected intense blue while nuclei in intact cells are oval

and surrounded by the pinkish-orange cytoplasm. Under certain conditions,

free nuclei tend to aggregate; such aggregation is reduced when a

minute amount of Triton-X 100 or Tween-80R is added, at a rate of

approximately 10-50 Il/liter saline. In the Biles (1973) hemacytometer,

nuclei present themselves as totally spherical particles measuring

approximately 2 ) in diameter. Both nuclei and intact cells sink to

the floor of the hemacytometer a few minutes after preparation. The

advantage of this method of determining hemolysis over conventional

photometric techniques is far greater simplicity and not having to be

concerned with an inherent contributor to the total variation caused by

fluctuations in bloodmeal size. Preliminary experimentation indicated

that hemolysis was variable from individual to individual, and thus

plotting of hemolysis rates of individuals would generate a frequency

distribution that could be analyzed and used in comparing treatments by

setting up profiles, where the bizarre elements are not lost in the

average. Furthermore, if hemolysis in extracts was measured photo-

metrically the small aliquots and low hemolysis rates would create

more technical difficulties, while if dog or human erythrocytes are

used, the exact number introduced would have to be determined; with

chick cells, one can sample from a large distribution of cells, as

in a clinical white cell differential count, or merely count all cells

and nuclei.

Extract preparation

Extracts of salivary glands (SG) and midguts (MG) were prepared

following the method of Day (1954), with some modifications to be

described. Salivary gland extract (SCE) was prepared by removing

pairs of salivary glands into saline in a high humidity chamber until

the required number was accumulated, when the saline was replaced with

distilled water; the rapid movement of fluid and particles spheruless)

from the gland cells into the saline was immediately evident, as described

by Day (1954). The first modification came with removal and holding

of this first extraction in a capillary, leaving only the SG, which are

subjected to coverslip squashing with insect forceps under the dissecting

scope; all cells are totally disintegrated, an edge of coverslip lifted

a mm from the slide and the original fluid re-introduced and mixed with

the squashed material and fluid. The slip is removed and the prepara-

tion placed at 4 C in a frost-free refrigerator, which dissipates the

preparation within a few hours, while inhibiting any microbial growth.

Just prior to use, the slide with extract is placed in a high humidity

chamber and saline added in the amount called for in the test protocol.

When MG are placed in distilled water, they swell within a

minute, and the epithelium peels away from the basement membrane as it

disintegrates, yielding a milky suspension inside the clear membrane.

Such sacs are sucked through the orifice of a capillary drawn to a size

that requires squeezing of the sac to get through it. After drawing

the sacs through the orifice three times, totally macerating the

material, the preparation is filtered by sucking it through a sterile

cotton plug capillary filter with 2, 10 pl distilled water washes,

yielding a clear fluid, which is henceforth processed as was SGE.

In addition to extracts of these organs, fresh coverslip squashes of

SG and MG were prepared to which were added microfilariae or chick red

cells, to support the results of extracts and to provide factors that

might be destroyed by extract preparation.

One other substance prepared for testing its effects on micro-

filarial vitality and hemolysis was blood meal extract, which was the

liquid portion of centrifuged blood meals. Ten fresh blood meals were

placed together in a drop of saline which was then drained; the meals

were teased apart after the epithelium was removed, and meals were

pushed down the cut end of a Pasteur pipette, the end of which was

sealed with heat fusion. Making the cut where the glass began its

flare facilitated the process,since this created a small funnel which

helped to get the material into the tube without loss. These were then

centrifuged at high speed for 15 minutes; the solid material was dis-

carded after separating the tube at the junction of liquid and solid

using a diamond pencil.

Additional Procedures and Materials

Inoculation per ano, also known as the "enema" technique, of Cx.

nigripalpus and other species with microfilariae and blood components

was accomplished using drawn capillaries, the ends of which were fire-

polished to facilitate penetration while avoiding rupture of the gut.

Tips were brought close to the flame of an alcohol lamp from below,

and, depending on the fineness of the draw, held a few mm from the flame

for a moment and examined for polishing under the scope. A brief

flicker of yellow color in the flame signaled that polishing had

commenced, but complete sealing can easily occur, in which case the

effort is simply renewed, until the proper result is attained. As

much as 4 ul of fluid could be introduced in this manner, although

2-3 ul was the amount most often delivered. Mosquitoes were prepared

for per ano procedure by attaching them with white glue to a slide by

the anterior thorax; the slide was inverted and placed at an angle so

that upon drying and re-inversion the tip of the abdomen projected

upward and slightly toward the operator. A scalpel edge applied to the

glued area easily dislodges treated mosquitoes into appropriate holding


Injection of the midgut blood meal with microfilariae was easily

accomplished using drawn capillaries; an abrupt draw gives the shaft

stability, as flexion is undesirable. As with clinical technique, the

best sign that the inoculating shaft is in proper position is the appear-

ance of cells in the shaft extremity; the outward pressure exerted by

the meal is usually sufficient for this to occur, although very light

suction is sometimes helpful. Before inoculating the fluid, a brief

moment is allowed to pass to permit hemolymph-sealing of the area.

Gradual delivery curtails the tendency of gut contents to leak when

pressure builds. Forceps tines applied next to each side of the

capillary shaft near the point of penetration reduces leakage after

withdrawal. Injection capillaries are more finely drawn and sharply

pointed than per ano capillaries.

Excised midguts were inoculated with microfilariae and blood

fractions at the cardiac sphincter, just posterior and adjacent to the

origins of the diverticulae. The needle for such a procedure is fire-

polished, as for enema, to facilitate injury-free penetration, which is

accomplished by grasping the edge of the sphincter with a forceps, and

inserting the needle throughthe sphincter down into the anterior midgut,

all the way into the far more expanded area of the posterior midgut,

which contains the secretary epithelium.


Buffers, anticoagulants and antibiotics were used in specific

experiments to condition mosquitoes prior to and during feeding; the

buffers and anticoagulants are described in the Results section. Anti-

biotics mixtures, in sugar or saline, depending on the experiment, were

prepared as follows: 50 pg each of neomycin sulphate (675 mcg/mg),

streptomycin sulphate, sodium penicillin G, and 100 pg nystatin (anti-

fungal antibiotic) per 100 ml saline or sucrose for preconditioning, or

for feeding antibiotic simultaneously with blood, 500 ig of each anti-

biotic is then mixed with 1 ml of the blood.

Per Os bloodfeeding via sugar-feeding mode

The feeding apparatus for this procedure is shown in Figure 11-7.

Several small 6 mm O.D. glass tubes each about 3 cm long with a 4 mm





28mm "
n 17|

..l ..... ._.j-uri
6mm O.D.I
Smm I.D. I 11Omm

"------------ ltOOm! --

67mm CUP

Figure 11-7. Special apparatus for blood feeding mosquitoes in
the sugar feeding mode-schematic.

inner diameter were charged with sugar solution containing microfilariae

or red cells. The end of the tube that projected into the cup was

ground, to allow even dispersion of the substance around the orifice;

about 100 yl could be placed in such tubes without affecting the important

operating characteristic, namely, the maintenance of a non-protruding

column of liquid in the tube. The purpose of this is to prevent the

mosquito from drinking in areas that are free of microfilariae or

cells; these settle to the bottom by gravity, and, if a bulge occurs,

clear areas around the upper edges of the protrusion can allow particle-

free imbibition. All such preparations are offered at room temperature

only. Mosquitoes to be so fed are starved for at least 24h before

feeding. The top of each tube projecting from the tube holder on top

of the cup is sealed with a cork to retard evaporation.

Wound Assay

The presence of injured microfilariae in the midguts of Cx.

nigripalpus was first reported by Nayar and Sauerman, (1975), although

the same condition was reported previously in other Culex and some

Anopheles by Coluzzi and Trabucchi (1968), using Dirofilaria reopens,

and in Cx. territans ingested Foleyella flexicauda (Benach and Crane,

1975). In the present study, microfilariae were separated from the

other blood meal components after hemolysis of red cells to facilitate

inspection, placed in droplets on a slide and allowed to dry. The

method of Awogun (1978) was used, with the modification that preliminary

fixing with acetic is not necessary, the addition of the Hoyer-fixative-

stain (hematoxylin) mixture serving the purpose quite adequately. The

great advantage of this technique is the simultaneous fixation, staining

and mounting, and the staining results are excellent for the purpose.

When large numbers of microfilariae were present in the preparation,

a sample count was taken to determine the occurrence and rate of

wounding, which, like vitality and hemolysis, was variable in Cx.

nigripalpus, and thus a group of mosquitoes generated a frequency

distribution of portion wounded.

Histological Examinations

Material for histological examination was prepared essentially as

described in Nayar and Sauerman (1975). Serial sections of 7 p slices

of groups undergoing a variety of treatments were prepared and examined

for differences between them and controls or other model mosquitoes,

regarding blood meal processing and other features that could help to

explain the vector-parasite relationships.


Thick smears of midguts were heat-fixed and stained in Loeffler's

methylene blue for bacteriological examination. While midguts were

usually devoid of bacteria or supported barely detectable populations,

in some cases large discrete boluses of bacteria were present, apparently

digesting the meconium. Antibiotic preconditioning was apparently

successful in destroying these organisms, if the smears can be taken as

a measure of activity (Micks and Ferguson, 1961), but culturing of

midguts was not done; thus, although digestive processes were disturbed

by administration of the antibiotic, as to whether it really eliminated

the flora was not determined. It seems safer to say that "suppression"


Thin smears of blood and other substances were usually fixed with

absolute methanol and stained with Giemsa 1:10 for 10 minutes and washed

in pH 7-5 phosphate buffer. This was adequate, although superior

results are obtained if the smear is stained for 30 minutes.

Data Presentation and Statistical Analysis

Counts of microfilariae, prelarvae, hemolyzed cells and vitality

levels from individual mosquitoes are highly variable and therefore

generate frequency distributions, when groups of such individuals are

plotted. In several instances, it was apparent that these distribu-

tions were not "normal" and that simple presentation of arithmetic means

and standard deviations would be inappropriate in terms of representing

central tendency and detecting shifts in individual composition of

particular regions of the distribution, and, therefore, two representa-

tions of the data are presented. One is the geometric mean, the anti-

logarithm of the arithmetic mean of log values of elemental data, and

geometric standard deviation. The basis of the normal standard devia-

tion is the squaring of deviant values, thereby conferring greater

weight on deviant values, and, since the average deviation decreases

with increases in sample size, logically places weight on sample size.

But, when data is heavily "skewed," as was at times the present case

due to large portions of 0 and low value counts, the few high values

pull the arithmetic mean disproportionately away from the "average" or

"representative" mosquito, the arithmetic mean being the most efficient

locater of central tendency when the distribution is normal (Langley,

1971). Use of the geometric mean, since it has its basis in logarithmic

transformation, reduces the weight of the extreme values and is more

representative of central tendency, which might otherwise be better

located by the median.

The geometric mean, when used with heavily skewed data, also, as

the above might imply, transforms the shape of the distribution from

skewed to normal, thereby fulfilling a requirement of parametric statis-

tics (normal-shaped distribution), thereby allowing the use of this

more powerful body of statistics. These methods of hypothesis testing

cannot be used with data that are heavily skewed or in any way non-

normal, and thus non-parametric methods, which are less efficient

(Sokal and Rohlf, 1969), would have been mandated without further

transformation. The parametric statistic used for hypothesis testing

in this study was the Student's 't' distribution, calculated on the

basis of different sample sizes (Steel and Torrio, 1960; Snedecor and

Cochran, 1968), where the sums of squared deviations are 'pooled' and

divided by the sum of sample N's 2, forming a variance that is used

to calculate 't' by dividing the difference between means by the sample

estimate of the standard error of that difference (calculated from the

pooled variance). Degrees of freedom are calculated from the combined

sample sizes 2.

The second representation of data is the "profile," which is

essentially the grouping of a frequency distribution into classes allows

a more meaningful visualization of shifts or dislocations possibly

associated with a treatment than ungrouped basic data. The establishment

of categories is arbitrary, as it is in all such cases, and each category

is not of the same proportional size in all profiles, since special

attention to the more deviant classes was of paramount importance re-

garding investigation of mechanisms. Means and standard deviations,

while excellent statistics for central tendency and dispersion, lost

sight of individuals and their position on a measurement continuum, and

different distributions can have the same means and same average squared

deviation. The profile, the categories of which are arbitrarily assigned

equivalent levels of susceptibility, allows visualization of the shifts

of individuals through categories associated with different treatments

without obfuscation from basic data clutter.

The relationship between number of microfilariae ingested and

number of prelarvae in the tubules was statistically analyzed using

conventional "least-squares" formulae. Correlation coefficients are

included as a measure of how well the linear regression line approximates

the data.

The following were used in calculations:

Geometric mean (X) and standard deviation (s)

X geometric = antilog (log E X/N) 1

s geometric = antilog E log X ( log X)

t = X X2 N ; s = 2 + x22

s1 2 NN2 N1 + N21 2*

(t-test statistic) (sample estimate of (pooled variance)
standard error of

Each x value was incremented by 1 prior to calculation of mean and
standard deviation of logarithmic values because of the presence the
0 values in the data. Therefore, the geometric values are actually
the antilog 1 of the mean and standard deviation.

** Degrees of freedom

Linear regression and correlation coefficient

slope = Z XY ZXIY
2 2X
.x2 .X)2

Y-intercept = SY (slope) ( X)

correlation coefficient = XY / /(SX2)(Y2)

Experimental Results and Discussion

Culex Susceptibility Patterns Background

To investigate the mechanisms that would explain susceptibility

level variation of the pattern found in Culex spp., Cx. nigripalpus

was chosen because the salient characteristics are more sharply defined

in this species than in Cx. quinquefasciatus (Nayar and Sauerman, 1975),

with greater portions of mosquitoes exhibiting the characteristic

immobilizing of ingested microfilariae and resultant low intensity of

infection; the additional potential mechanisms of rapid coagulation of

the bloodmeal and hemolysis of red cells with crystal formation are

likewise more extreme than in the other Culex described in the literature

with the possible exception of Cx. territans, all of which implies that

experimentally-induced dislocations will be more rapidly and accurately

detected. Furthermore, this insect is readily available year round in

the region in which this study was conducted and has been implicated in

the transmission of St. Louis encephalitis, turkey malaria and dog

heartworm (Nayar, 1980). It is the aim of this study to explore factors

that induce entrapment and loss of microfilarial vitality utilizing and

extending the addition:deletion strategy of Kartman (1953a), who essentially

was adding the "missing" salivary gland anticoagulant. Entrapment is

visualized as a function of absence of anticoagulant secretion by the

salivary glands, as first suggested by Kartman (1953a), or the presence

of midgut coagulants, which have been described in mosquitoes (Buck,

1957). Microfilarial vitality loss presents a more complex problem,

with possible sources including the presence of antiblastic, toxic

factors from host or host-flora (which may indirectly depend on movement

hindrance for their effect), and the traumatizing of microfilariae by

pharyngeal valve spines or needle-sharp oxyhemaglobin crystals, them-

selves the product of hemolysis due to host factors or flora metobolites.

The Effects of Hindrance of Migration of Microfilariae to the Malpighian


Although the increase in migration rate of D. immitis microfilariae

to the Malpighian tubules in Ae. aegypti, bloodfed via membrane on

heparinized, infectious dog blood, was not ultimately linked to a change

in either "host efficiency" or "infective potential," the significant

fact remains that Kartman (1953a) had succeeded in experimentally

altering at least one phase of the Ae. ao.,ypti:D. immitia vector-parasite

relationship. The experiment was designed to test a hypothesis, based

on his observation that the coagulation time of the ingested bloodmeal

is a salient variable distinguishing the highly susceptible An.

quadrimaculatus from the far less susceptible Ae. aegypti and Cx.

quinquefasciatus, that the speed of migration, or conversely, the

hindrance of movement, of ingested microfilariae through the midgut to

the target tissue determined the susceptibility level, possibly by in-

creasing or decreasing the exposure time to salivary gland or midgut

secretions that were antiblastic with regard to the microfilariae. The

ingenius tactic of "adding" an absent factor, in this case, salivary

gland anticoagulant, to simulate the susceptible mosquito, to determine

if the variable factor mediated the susceptibility level, was a hallmark

in the investigation of filarial host-parasite relationships.

However, although Kartman (1953a) had also determined that the

pattern of infection in his Culex group strongly suggested digestive

process involvement, he had not subjected representatives of the genus

at his disposal to the same protocol used with AP, ae2gyti, and,therefore,

this seemed an appropriate starting point for the present investigation.

Kartman (1953a) had shown, using Cx. quinquefasciatus, and this had been

corroborated by Nayar and Sauerman (1975), using Cx. nigripalpus,

that the characteristic poor host efficiency, defined as the number of

target tissue (Malpighian tubulo) developing forms divided by the total

number microfilariae ingested, was a function of low number of micro-

filariae reaching the tubules, that the "barrier," manifest in the midgut,

lay somewhere prior to the target tissue, since there was ample evidence

in these species that there were no further barriers of significance

once the target tissue had been reached. This contrasts sharply with

the pattern exhibited by the Ae. aegypti:D. immitis vector-parasite

relationship, where the major barrier, involving arrest of development,

awaits the microfilariae in the target tissue (Kartman, 1953a; Nayar

and Sauerman, 1975). The possibility that Kartman's hypothesis might

apply to Culex, where the prolonged midgut experience, resulting from

being trapped in the clot, could either have sapped the strength of the

microfilariae or subjected the microfilariae to contact with a substance

which was with little effect when exposure was brief, was therefore

sizable. In addition, Cx. guinquefasciatus and Cx. nigripalpus vector-

parasite relationships with D. imnitis resemble one another with respect

to rapid bloodmeal coagulation, hemolysis of erythrocytes and crystal

formation (Nayar and Sauerman, 1975). Therefore, the hypothesis was

tested using the basic addition:deletion strategy initiated by Kartman

(1953a) and expanded in this case, to uncover the mechanisms responsible

for the low susceptibility level.

In contrast with Kartman's (1953a) approach, no attempt here was

made to measure the actual speed of migration of microfilariae through

the midgut to the Malpighian tubules, because it had not yet been shown

to induce a significant dislocation in the vector-parasite relationship,

as measured by an increase in host efficiency or some other susceptibility

parameter. Kartman, for example, had discontinued his anticoagulant

experiments after the 24 measurement and therefore did not determine

if the increased rate was associated with an increase in developing

forms, although his own data indicated that this would be the appropriate

parameter to measure, as a huge portion of microfilariae that attain

the target tissue do not develop further, as has since been noted by

Nayar and Sauerman (1975), who discussed the tubule barrier, and, using

Brugia spp., by Beckett and MacDonald (1971) and Schacher and Khalil

(1968), where arrest occurs in the thoracic musculature. However, for

Cx. nigripalpus, the "host efficiency" parameter of Kartman (1953a) is

ideal, since, as he puts it (1955b), "It is of interest to note that

although Cx. pipiens was an unfavorable host for survivorship of worms,

if they did survive then it was a favorable host for their growth" p.1068.

Therefore, the number of prelarvae in the tubules relative to the number

of ingested microfilariae (host efficiency) is a sufficient parameter to

detect dislocations from the normal vector-parasite relationship

induced by the addition:deletion strategy. An. quadrimaculatus controls

were fed simultaneously to insure that the absence of response was not

due to loss of microfilarial vitality from the treatment, and the

bloodfed + 24h assay time avoids having to contend with the additional

variable of mosquito mortality associated with this species (Weiner and

Bradley, 1970; Nayar and Sauerman, 1975; Hamilton, 1975).

One extremely important aspect of most filarial:arthropod relation-

ships in terms of delineation of mechanisms is the individual variation

in susceptibility level; although the existence of just such variation

is discernable from the results of Kartnan (1953a) and Nayar and

Sauerman (1975), while it has been alluded to in much earlier work

(Hu, 1951; Yen, 1958). One is usually confronted with an array of host

efficiencies, developing form ratios and infective potentials, with the

frequency of any particular level of susceptibility being characteristic

for the species as a whole. For example, if even one individual of

thousands dissected is positive for infective stages, the species cannot

be categorized as totally refractory, and, on the other hand, if 5%

of the individuals of a species are refractory to infection, the species

cannot be labelled totally susceptible.

It is apparent from the work of Kartman (1953a) and Nayar and

Sauerman (1975) that the usefulness of the number of ingested micro-

filariae as an estimator of future events, such as host efficiency,

developing form ratios and the like, is variable depending on the species.

While in a highly susceptible species such as An. quadrimaculatus, the

number of microfilariae ingested is tightly correlated with the number

of prelarvae (microfilariae in the Malpighian tubules) and each successive

stage (with a decrement constant factor applied) found, this is plainly

not so with either Culex spp. or Ae. aegypti. For example, 60/o of Cx.

quinquefasciatus ingesting 576 + 42 microfilariae of D. immitis had only

2-6 prelarvae per female, but some had none and others had more (Nayar

and Sauerman, 1975); thus, the ingestion of more than 100 microfilariae

usually yields few if any prelarvae, (thereby setting the pattern of

low susceptibility by which both species are characterized) but on

occasion can yield many prelarvae, far outside of the "general" pattern.

Experiments and results

To confirm that the number of microfilariae ingested by Cx.

niripalpus is indeed a poor estimator of prelarval load or "host

efficiency," groups of mosquitoes were fed on the infectious dog; after

24 the midgut and tubules were removed, separated and counts made of

midgut microfilariae and tubule prelarvae (Table IL-l1 The shallow slope

and weak correlation coefficient indicate that there is little relation-

ship between the number ingested and the number of prelarvae, which is

intuitively obvious given that most values of ingested microfilariae

are associated with low to zero values of prelarvae. Similarly, the

occasional highly susceptible individuals are not the result of a

particularly high or low number of microfilarlae ingested. Since any

proposed mechanism must explain the existence of various individual

host efficiencies and how they are generated, it is clear that the

mechanism is not merely the ingestion of large numbers of microfilariae

producing an array of prelarval loads.

Before initiating experiments using the addition:deletion strategy,

the basic frequency distribution of individual susceptibility generated

by normal (untreated) Cx. nigripalpus feeding on the infectious dog and

dissected no less than 24 later for prelarvae in the tubules, was

determined (Table 11-2 Less than 5;. are "susceptible" by An.

quadrimaculatus standards, and a tenth covers the percentage of individu-

als carrying more than 10 prelarvae. The predominant group, as expected

based on previous work with this mosquito (Nayar and Sauerman, 1975),

accounting for 90% of this sample, is comprised of individuals harboring

10 or less microfilariae (Table n-2) and it is this group upon which the

general susceptibility level of the species is based.

To test the hypothesis, first advanced by Kartman (1953a), that

hindrance of movement of microfilariae through the midgut of the mosquito,

caused by absence of salivary gland anticoagulant, might account for the

observed low susceptibility level, his strategy, that is, addition of

the variable factor to simulate the susceptible mosquito, was followed.

Three anticoagulants--citrate, heparin, and i3TA (ethylenediamine

tetraacetio acid) -were tested, by feeding 7-14 day old Cx. nigripalpus

via the membrane feeder. The distributions generated by this "addition"

protocol did not differ significantly from that seen in the "natural"

group (Table 1-2)and, if the purpose of the addition was to simulate the

An. Quadrimaculatus vector-parasite relationship, it can be safely said

that this was not achieved. Thus, either movement hindrance and entrap-

ment were not factors playing a role in determining the characteristic

susceptibility level, or the simulation of the anticoagulated An.

quadrimaculatus bloodmeal had not occurred using the protocol. The

blood in the membrane feeder was fluid throughout the feeding interval

(2-3 with a majority usually having fed during the first half hour),

and to determine if the same state was maintained after entering the

mosquito, midguts were dissected from mosquitoes within 15 minutes

after imbibition of blood containing e;'ch anticoagulant. Regardless

of the anticoagulant, approximately 95, of all meals were clotted to

various degrees, indicating that the itico 'uluit had failed to

function as expected (Table II-3). Con iderin~ the difference in modes

of action -both calcium chelation and binding ( iDTA and citrate)

and thrombin formation inhibition (heparin) (Simmons, 1968), this was

quite provocative, and strongly implied the presence and activity of a

coagulant which induced clotting even in anticoagulant-treated blood -

the presence of coagulants in mosquitoes was reported by Buck (1937).

A second implication is that the salivary gland anticoagulant activity

level cannot be assayed in directly bloodfed Cx. nigripalpus, although

the absence of such activity might easily be inferred from the compara-

tive clotting times of An. luadr.lmaculatuo and Cx. nigripalpus (Nayar

and Sauerman, 1975) fed directly on hosts, and, on a generic basis,

from the absence of activity having been reported in Cx. salinarius,

Cx. quinquefasciatus and Cx. restu ns (etc alf, 1 '), and Cx. pipiens

(Yorke and Macfie, 1924), while low level activity was observed in

Cx. pipiens (Buck, 1937; Hudson, 1914) and Cx. tarsalis (Hudson,

1964). Since no direct infonnation on either salivary gland anticoagulant

activity or midgut coagulants activity existed for Cx. niripalpus,

determinations were made for both entities. The clotting time of blood

was clearly increased over controls (Table 11-4), although this effect

would not be measurable in vivo, since clotting of bloodmeals in the

midgut is nearly instantaneous (Table II-3).

The discovery of salivary gland anticoagulant increased sub-

stantially the likelihood that the rapid coagulation of ingested dog

blood and ensuing hindrance of movement of D. immitis microfilariae

was a function of the presence of a coagulant rather than absence of

anticoagulant, and a characterization of this component was therefore

necessitated. A stomach secretion from Culiseta (ETheobaldia) annulata

caused rapid coagulation of blood, even when the blood was treated with

Anopheles salivary gland secretions ( Tuck, 1957); the secretion was

not present in the anterior midgut or in male midguts. A coagulin was

present in Cx. pipiens, but in much smaller concentration.

The initial assay tested the extract from 10 Cx. nigripalpus

female midguts mixed with 50 pl dog blood treated with 100 units sodium

heparin per ml, 5 times the amount needed to maintain the blood in the

liquid state indefinitely. In all trials, the blood coagulated within

1 minute, with first threads detected within 30 seconds, almost

instantaneously (Table 11-5). 'hen the anticoagulant level was raised

to 1000 units heparin/ml, again, rapid coagulation of the sample was

observed. No effects on clotting times were observed using either

An. quadrimaculatus or Ae. aegypti extract, even when 10 midguts were

tested on blood treated with 10 units heparin/ml do6 blood, which is

right on the verge of clotting srontaneou ily. The coagulant effect

did not wane until 0.5-1.0 midguts were mixcd with dog blood treated

with 1000 units of heparin/ml (Table I1-5).


As Buck's (1937) results are not accompanied with quantitative

data, a comparison of the strength of the coagulins in Cx. nigripalpus

with those in Cu. annulata or Cx. pipiens is not possible, but the mode

of action seems similar. The results here attest tothe presence of a

powerful coagulin, similar in strength to Culiseta and greater than that

of Cx. pipiens, with a surprisingly versatile chemical property, in that

it can override the anticoagulant properties of substances with

different modes of action. The mode of action of this coagulant remains

to be determined, and it cannot be stated that it induces or is involved

with the normal "cascade" of coagulation. There were indications that

citrate could prevent the effect when used in higher concentrations,

but in vivo tests were frustrated by refusnl of mosquitoes to feed on

blood thusly treated.

Thus, the addition of anticoagulants, in the levels used here,

failed to satisfy the requirement of the addition:deletion strategy,

namely that the equivalent of an An. nuadrimaculatus bloodmeal, liquid

in state and offering little hindrance to trophic movement by micro-

filariae, be produced in Cx. nigripalpus; the meal did not remain

liquid, a clot being formed as in "normal" mosquitoes, and movement

hindrance from coagulation was not prevented. The basic vector-parasite

relationship, as measured by the prelarvae per female frequency distri-

bution, remained essentially unchanged. Although the proper biochemical

characterization of the coagulant is beyond the aims and scope of this

investigation, 2 lines of evidence suggest that it has a protein com-

ponent: de Buck's (1937) characterization of the Culiseta coagulant-

unaffected by prolonged drying, heating to 990C, heating of the dissolved

secretion at 4000 for 30 minutes; dissolved secretion deactivated by

500C for 15 minutes suggests this, and, in the present work, the use

of glass-bead defibrinating of blood, which satisfied the requirement

that a low-hindrance bloodmeal be induced in Cx. nigripalpus, supports

a contention that fibrinogen is the protein involved. Later, tests on

serum collected from blood incubated in thrombinized vials for 1 at

37C (Sigma Chemical Co. Bull. No. 850, 1977) proved negative for clotting.

Initially, the use of the glass-boad defibrinating process had

been rejected in favor of the addition of exogenous anticoagulant for

2 reasons: 1.) the purpose of the protocol was to simulate, as nearly

as possible, the An. quadrimaculatus midgut environment, and, while

there may be little chemical affinity between a polysaccharide such as

heparin and proteinaceous salivary gland anticoagulants, the mechanism

whereby the coagulation cascade is interfered with is essentially

similar and the net result, the absence of fibrin formation, is the

same, while glass-bead defibrinating does not interfere with the

cascade, since it merely prevents clotting by disallowing continuous

fibrin strand formation. 2.) The very procedure of glass-bead de-

fibrinating, where the blood in churnede" with beads for 30 minutes,

seemed more likely to traumatize microfilariae. However, with the

failure of the anticoagulants to function as required, the testing and

use of glass-bead procedure was mandated. Saponized droplets of blood

samples taken following churning were examined for moribund or abnormal

microfilariae but none were seen, and Giemsa-stained microfilariae on

thin smears were not visibly wounded, indicating that the preparation

was suitable for use in the membrane feeder.

In the interests of volumetric equivalency, the dilution factor of

0.1 ml of 0.9% NaC1, used in all anticoagulant experiments to dissolve

the anticoagulant prior to mixing with blood, was also added to the glass-

bead defibrinated blood. While approximately the same number of micro-

filariae were ingested per female as in the anticoagulant-treated

groups (Table 11-2), more than twice as many mosquitoes had An.

quadrimaculatus-level prelarval loads (Table 11-2), and the portion of

totally refractory mosquitoes decreased dramatically. Nearly 30% of the

total number of mosquitoes fed on glass bead defibrinated blood had

prelarval loads in the "moderate" susceptibility level category (Table

11-2). A similar but smaller-sized shift of the distribution from the

refractory and low susceptible categories to the moderate-but not

upper-level categories had been observed in anticoagulant-treated

groups but there was a greater than 5%/ chance that such a group could

have come from the same population as the "natural," directly fed

group, while the glass bead defibrinated group differed significantly

(p <5%) from the directly-fed group (Table 11-2).


The above experiment gives "qualified" support to the hypothesis

first advanced by Kartman (1953a) and supported by later work (Nayar

and Sauerman, 1975), that the hindrance of microfilarial movement is

a factor in determining the susceptibility level of mosquitoes to

filarial infections. The mechanism for this hindrance in Cx. nifripalpus

is the rapid clotting of the ingested bloodmeal, due to the action of

a powerful midgut coagulin, whose net effect was circumvented using a

physical technique. However, the mechanism accounts for only a portion

of the poor host efficiency characteristic of Cx. nigripalnus. The

mean prelarvae per mosquito, while differing significantly from the

mean of the "natural" group, implying that the treatment produced a

real effect, was nevertheless still only in the moderate range. Thus,

while the qualities of the An. qualrimaculatus bloodmeal had been "added"

to the Cx. nigripalpus:D. immitis vector-parasite relationship, the

susceptibility level of this sample's population, while greater than

that of the "natural" group, was far lower than that of the "susceptible"

model, An. ouadrimaculatus, implying that either other aspects of the


hindrance factor were not being ardresned by the design or that other

mechanisms were operating in addition to those associated with movement


The Relationship of Hemolysis and Crystal Formation to Susceptibility
of Cx. nigripalpus to Infection by D. immitis


Another distinguishing feature of the Cx. nigripalpus:D. immitis

vector-parasite relationship is the hemolysis of ingested red blood

cells and subsequent production of needle-shaped oxyhemaglobin crystals

in the bloodmeal (Nayar and Sauerman, 1975). Hemolysis of erythrocytes

from Foleyella infected frogs occurred in the midguts of Cx. territans

within 20 min. after ingestion (Benach and Crans, 1975), leading the

authors to state lysiss of red blood cells and microfilarial death

shortly after ingestion suggests barriers related to salivary gland or

midgut secretion or to the mechanical action of the bucco-pharyngeal

armature." With Cx. pipiens, three types of secretions follow blood

feeding (Arnal, 1950), and it was concluded that lysis of red cells was

the result of metabolites or enzymes from bacteria that penetrated the

meal. It is unlikely that hemolysis is the result of midgut digestive-

proteolytic activity, which peaks at 24h after feeding (Gooding, 1972),

and protease activity graphs (Briegel and Lea, 1975) show little or no

activity during the first hour post-blood feedings, long after hemolysis

has occurred. This does not rule out the possibility of a hemolytic

agent potent at extremely small concentrations being present. However,,

no hemolysins have ever been detected in mosquito salivary gland prepara-

tions, even when specifically assayed for it (Gooding, 1972).

In preliminary experiments, crystals were observed even in the

bloodmeals of Cx. nigripalpus membrane-fed on glass bead-defibrinated

dog blood (GBDF) and anticoagulant-treated dog blood, and the possibility

that they could, through their presence or absence, mediate the

susceptibility level, through impedence of microfilarial migration or

inflicting of wounds, made mandatory a more thorough examination and

characterization of the phenomenon. The released hemaglobin itself

radically changes the chemical environment in which the ingested

microfilaria finds itself.

Results and Discussion

To determine whether the crystals were involved in mediating the

vector-parasite relationships, crystals were deleted by membrane-

feeding Cx. nigripalpus females, along with An. quadrimaculatus vitality

controls, on anticoagulant-treated chicken blood, to which microfilariae,

separated from an equal volume of infectious dog blood, had been added.

Since the addition of anticoagulant to dog blood had not demonstrably

affected the vector-parasite relationship in general (Table 11-2), or

coagulation time in particular (Table II-3), and, since the coagulation

time of chicken blood in Cx. nigripalpus is even faster than dog blood

(Nayar and Sauerman, 1975), any increase in the portion of susceptibles

or even moderate susceptibles could bo attributed to either the absence

of crystals, as chicken hemaglobin does not crystallize, at least not

so as to produce observable entities under light microscopy within the

time frame considered relevant here, or to the other variables attributable

to differences in blood composition.

Although there is an indication of come microfilarial vitality loss

as seen in the An. cuadrimaculatus control group (Table II-2) (most likely

due to the use of the distilled water microfilarial separation method,

later improved upon by using saponin), there was no indication whatever

that the absence of crystals improved the susceptibility profile

(Table II-2); the same large portion of totally refractory and low

susceptible mosquitoes dominates the profile. While these results are

not, by themselves, conclusive, because of the possibility of un-

controlled variables from the microfilarial separation procedure or

blood composition that could differentially affect the reaction in

Cx. nigripalpus compared with An. quad rim-cculatuq, they do not support

the lethal-crystal formation hypothesis. However, an interesting

phenomenon was observed during the course of bloodneal examination for

crystals; instead of the usual picture seen with dog blood, where one

observes red cells, red fluid and crystals, suspension of the chicken

bloodmeal in normal saline revealed the presence of large numbers of

2 p diameter spheroids. That these were indeed chick red cell nuclei

(the erythrocytes of bird blood are nucleated) was confirmed by ob-

serving the hemolysis of 3X washed chicken erythrocytes under a coverslip,

running distilled water (later saponin) under one side of the slip with

a fine capillary. Apparently, the nuclei are not at all subject to

hemolysis, by either distilled water or saponins, since 50 pl of washed

cells in 15 ml distilled water, followed by centrifugation, produces

a pellet of nuclei. This accounts for the observation that hemolysis of

chicken blood does not produce the clear, transparent red solution seen

with dog or human blood; instead, a cloudy, only slightly less opague

suspension is seen and clear fluid is produced only by centrifugation.

The significance of this phenomenon was that hemolysis had occurred and

had left a marker other than released hemaglobin, that this marker was a

measurable entity without having to run a laborious procedure, and

without the need to calibrate for bloodmeal size.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs