Isoprenylation of testicular cell proteins in rat seminiferous epithelium


Material Information

Isoprenylation of testicular cell proteins in rat seminiferous epithelium
Physical Description:
xiii, 184 leaves : ill. ; 29 cm.
Dugan, Jan Marie, 1965-
Publication Date:


Subjects / Keywords:
Protein Isoprenylation   ( mesh )
Testis -- chemistry   ( mesh )
Seminiferous Epithelium -- chemistry   ( mesh )
Spermatocytes -- chemistry   ( mesh )
Spermatids -- chemistry   ( mesh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1993.
Includes bibliographical references (leaves 175-183).
Statement of Responsibility:
by Jan Marie Dugan.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 49676489
System ID:

Full Text







This dissertation is dedicated to my parents, Barbara and

John Dugan, and to my fiance, Dwight Pitcairn.


I am extremely grateful for the invaluable guidance and

untiring support that my mentor, Dr. Charles Allen, has

provided for me. Through him I have learned the theory of

experimental design, the fundamentals of biochemistry and the

art of creative authorship. I am appreciative of the

assistance and productive criticism from my supervisory

committee members, Dr. Daniel Purich, Dr. Harry Nick, Dr.

Susan Frost, Dr. Chris West, and Dr. Don Cameron. Their

encouragement and advice were essential for the completion of

my projects. I am especially grateful to Dr. Cameron for

allowing me to visit his laboratory, for teaching me

testicular cell separation techniques and for his guidance and


I am also indepted to the faculty, staff and students of

the Department of Biochemistry for their friendship, their

helpful ideas and their support. In particular, I would like

to thank Mary Handlogten for my early training, Dr. Nancy

Denslow for sharing her expertise on many occasions, and Angie

Carter for her secretarial help. Finally, I would also like to

thank Dr. Ben Dunn and Dr. Brian Cain who made me feel

especially welcome in their laboratories.



LIST OF TABLES . ... .. vi

LIST OF FIGURES . . .. vii


ABSTRACT . .. ... xii


Background . . 1
Significance . . ... 17
Objectives . . .. 19


Introduction . . ... 21
Materials and Methods . .. 23
Results . . ... .. 32
Discussion . . 72


Introduction . . .. 84
Materials and Methods . .. 88
Results . ... .. 92
Discussion .. . 145


PROCEDURE . . .. .. 168


REFERENCE LIST. ... . ... 175



Table 2-1 Approximate Kms (AM) of Protein, Peptide and
Allylic Substrates for PFT, PGGT-I and PGGT-II 41

Table 2-2 Age Study of Prenylation of Recombinant
Proteins . .. .... 57

Table 3-1 Age-Dependent Ratios of Geranylgeranyl to
Farnesyl Incorporated into Proteins of the
Seminiferous Epithelium . .. 110

Table 3-2 Comparison of Age-Dependent Polyprenol
Incorporation into Proteins between Germ Cells and
the Seminiferous Epithelium .. 113

Table 3-3 Comparison of Polyprenol Incorporation into
Proteins between Pachytene Spermatocytes and Round
Spermatids . . 125

Table 3-4 Comparison of [3H]-Mevalonic Acid
Incorporation into Cholesterol and Dolichol between
Pachytene Spermatocytes and Round Spermatids .142


Figure 1-1 The Isoprenoid Biosynthetic Pathway 9

Figure 2-1 Cytosolic Protein Dependence for PFT and
PGGT-I Activities . ... 30

Figure 2-2 Cytosolic Localization for PFT and PGGT-I
Activities . .... 35

Figure 2-3 Peptide and Allylic Substrate Concentration
Dependence for PFT Activity . .. 38

Figure 2-4 Peptide and Allylic Substrate Concentration
Dependence for PGGT-I Activity .. 40

Figure 2-5 Time Dependence for PFT and PGGT-I .... .44

Figure 2-6 Peptide Specificity of PFT and PGGT-I 46

Figure 2-7 Inhibition of Farnesyltion of Bt-KTKCVIS
with N-acety-KTKCVIS . ... 48

Figure 2-8 Prepuberal Age Study of PFT Activity
Comparing Bt-KTKCVIS Peptide with p21"-r" Protein
as Substrates . . ... 51

Figure 2-9 Prepuberal Changes in PFT Activity in
Comparison with Adult Activity .. .... 54

Figure 2-10 Age-Dependent Changes in Protein PFT and
PGGT-I Activities in Spermatogenic Cells .. 56

Figure 2-11 PFT Activity in Spermatogenic Cells and
Seminiferous Epithelium . ... .61

Figure 2-12 Age-Dependent Changes of Spermatogenic
Cell Population and PFT Activity ... .64

Figure 2-13 PFT Activity in Isolated Pachytene
Spermatocytes and Round Spermatids in Adult and
23 Day Old Rats . ... 67


Figure 2-14 PGGT-I Activity in Isolated Pachytene
Spermatocytes and Round Spermatids in Adult and
23 Day Old Rats. ... . .69

Figure 2-15 PFT Activity in Pooled Sta Put Fractions
of Spermatogenic Cells from 23 Day Old Rats 71

Figure 3-1 One-Dimensional SDS-PAGE of [3H]-Mevalonic
Acid Labelled Seminiferous Epithelial Proteins 97

Figure 3-2 One-Dimensional SDS-PAGE of [3H]-Mevalonic
Acid Labelled Proteins from Subcellular Fractions
of Seminiferous Epithelial Cells .. 100

Figure 3-3 Two-Dimensional Gel Electrophoresis of [3H]-
Mevalonic Acid Labelled Seminiferous Epithelial
Proteins . . 103

Figure 3-4 Thin Layer Chromatography of Polyprenols
Released from [3H]-Mevalonic Acid Labelled
Seminiferous Epithelial Proteins .. 106

Figure 3-5 Polyprenol Analysis of [3H]-Mevalonic Acid
Labelled Proteins from Molecular Weight Regions
of One-Dimensional SDS-PAGE . ... .108

Figure 3-6 Age-Dependent Differences in Polyprenols
Incorporated into [3H]-Mevalonic Acid Labelled
Proteins of the Seminiferous Epithelium 112

Figure 3-7 Exogenous Mevalonic Acid Concentration
Dependence of Polyprenol Incorporation into
Proteins of 9 and 23 Day Old Seminiferous
Epithelium . ... 116

Figure 3-8 Exogenous Mevalonic Acid Concentration
Dependence of Geranylgeranyl to Farnesyl Ratios
Incorporated into Proteins of 9 and 23 Day Old
Seminiferous Epithelium . 119

Figure 3-9 Thin Layer Chromatography Profile of
Polyprenols Released from [3H]-Mevalonic Acid
Labelled Protein of Pachytene Spermatocytes and
Round Spermatids . .. 121

Figure 3-10 Spermatogenic Cell Type-Dependent
Differences in the Ratio of Geranylgeranyl to
Farnesyl Incorporated into [3H]-Mevalonic Acid
Labelled Proteins . .. 124

Figure 3-11 Age-Dependent Differences in One-
Dimensional SDS-PAGE Patterns of [3H]-Mevalonic


Acid Labelled Seminiferous Epithelial Proteins .

Figure 3-12 Age-Dependent Decrease in Specific
Activity of [3H]-Mevalonic Acid Labelled
Seminiferous Epithelial Proteins, Quantitated by
Densitometry . . .. 130

Figure 3-13 Similar Levels of Incorporation of [3H]-
Mevalonic Acid into Cholesterol and Dolichol with
Prepuberal Age . ... 132

Figure 3-14 Age-Dependent Differences in the Amount
of Label in Specific Molecular Weight Regions of
SDS-PAGE of [3H]-Mevalonic Acid Labelled Proteins
of the Seminiferous Epithelium ... .134

Figure 3-15 Age-Dependent Differences in Two-
Dimensional Gel Electrophoresis Patterns of [3H]-
Mevalonic Acid Labelled Seminiferous Epithelial
Proteins . ... .. 137

Figure 3-16 Spermatogenic Cell Differences in the
Amount of Label in Specific Molecular Weight
Regions of an SDS-PAGE of [3H]-Mevalonic Acid
Labelled Proteins . ... 140

Figure 3-17 Spermatogenic Cell Type Differences in
Two-Dimensional Gel Electrophoresis Patterns of
[3H]-Mevalonic Acid Labelled Proteins 144

Figure A Diagram of the Sta Put Cell Separation
Procedure . . ... .. .170

Figure B Electrophoretic Profile of Farnesylated and
Geranylgeranylated Proteins in the Seminiferous
Epithelium . .... .174
























bovine serum albumin



degree centigrade


disintegrations per minute


isopentenyl diphosphate

farnesyl diphosphate

geranylgeranyl diphosphate


3-hydroxy-3-methyl-glutaryl coenzyme A










phosphate buffered saline


PMSF phenylmethylsulfonyl fluoride

PFT protein farnesyl transferase

PGGT-I protein geranylgeranyl transferase-I

PGGT-II protein geranylgeranyl transferase-II

TLC thin layer chromatography

Tris Tris-(hydroxymethyl)aminomethane

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



Jan Marie Dugan

August 1993

Chairperson: Charles M. Allen
Major Department: Biochemistry and Molecular Biology

The isoprenylation of testicular cell proteins was

examined in prepuberal rat seminiferous epithelium,

spermatogenic cells, and in isolated pachytene spermatocytes

(PS, pre-meiotic) and round spermatids (RT, post-meiotic).

Studies show protein farnesyl transferase (PFT) activity in

the testes and spermatogenic cells of 23-day-old rats to be

2- to 3-fold higher than in 9-day-old animals and higher

than in older animals. PS and RT from adult rats had the

same level of PFT activity. PFT activity in PS from 23-day-

old rats was 2 fold higher than adult PS and RT, yet the

whole spermatogenic cell population from the same 23-day-old

rats showed even higher activity (4 fold). PFT activity

assayed in pooled Sta Put fractions of 23-day-old

spermatogenic cells was highest in the fraction containing


cells slightly smaller than pachytene spermatocytes,

possibly including secondary spermatocytes. Although the PS

are the most prevalent cell type at 23 days, they alone do

not appear to be responsible for the peak in PFT activity.

The seminiferous epithelium from different aged animals was

labelled with [3H]-mevalonate (MVA) and the protein bound

polyprenol analyzed following methyl iodide treatment. TLC

revealed geranylgeraniol (GG) and farnesol (F) as the only

products with GG/F ratios that decrease approximately 1.5

fold from 9 to 17 days with little change in older animals.

Polyprenol analysis of PS and RT from adult animals showed

GG/F ratios of 3.3 and 0.8, respectively, whereas the whole

spermatogenic cell population had a ratio of 1.9. Two-

dimensional PAGE of [3H]-MVA labelled proteins showed age

dependent changes of the levels of prenylation of at least

14 proteins. These results show cell type differences in

geranylgeranylation and farnesylation and suggest age and

cell dependent changes in protein acceptors in rat testes.




With the advent of post-translational modification of

proteins with isoprenoids and the implications of a role for

this feature in cell cycling came the hypothesis that

protein prenylation was involved in development. This work

describes changing features of protein prenylation during



Spermatogenesis entails many morphological and

biochemical changes for the differentiating germ cells

within the seminiferous tubules of the testes. Each germ

cell undergoes three main phases of spermatogenesis,

beginning with spermatogonial renewal and proliferation.

This phase involves six mitotic divisions. Type A

spermatogonia divide and differentiate into intermediate

spermatogonia and type B spermatogonia. The last mitosis of

type B spermatogonia results in the formation of

preleptotene primary spermatocytes. These cells migrate from

the basal compartment to the adluminal side of the tight


junctions of the epithelium formed by the Sertoli cells, and

replicate their DNA before entering the second phase of

spermatogenesis, meiosis. Distinct stages of meiotic

prophase are termed leptotene, zygotene, pachytene,

diplotene, and finally diakinesis to form secondary

spermatocytes. Immediately following diakinesis, a second

meiotic division occurs forming the haploid spermatids.

Spermiogenesis, the final phase of spermatogenesis, consists

of complex morphological transformations of the nucleus and

acrosome. At spermiation, the cytoplasm separates from the

flagellum, creating a residual body which is taken up by the

Sertoli cell in phagocytosis. The late spermatid is released

into the lumen of the seminiferous tubule.

The stages of spermatogenesis follow one another in a

regular fashion along the seminiferous tubule. At any point

in the tubule, there are well defined cell associations with

one to two generations of spermatogonia, spermatocytes, and

spermatids in each stage. Leblond and Clermont defined 14

separate stages of distinct durations that cycle in the rat

(1). Man has six stages, whereas the monkey and the mouse

have 12. The time interval between the reappearance of one

particular cell association stage is the cycle of the

seminiferous epithelium, and the succession of the 14 stages

along the length of the tubule is the wave. The wave repeats

itself along the length of an individual tubule in a

sequence of about 12 waves per tubule. In the postnatal rat,


the cycle begins with gonocytes appearing at day four of age

and ends with the first spermatozoa at day 45. In each

segment of the tubule the cycle repeats itself every 12 days

to provide a continuous wave of spermatozoa production

throughout adult life.

Sertoli Cells

The seminiferous epithelium consists of Sertoli cells

and germ cells organized in highly structured tubules by the

surrounding peritubular cells. In 1865, the Italian

physiologist Enrico Sertoli first identified the somatic

component of the seminiferous epithelium. He described the

cells as tall and columnar, extending perpendicularly from

the basement membrane to the lumen of the seminiferous

tubule, and enveloping the many associated germ cells (2).

These Sertoli cells perform a nurturing role for the

spermatocytes, analogous to the granulosa cells function

during maturation of the oocyte in the ovary. Tight

junctions, formed between the Sertoli cells, create a blood-

testes barrier that isolates and immunologically protects

the germ cells from the general circulation. The transport

and function of nutrients and hormones from the circulation

are mediated by the Sertoli cells in maintenance of germ

cell development. The Sertoli cells are the primary targets

for pituitary follicle-stimulating hormone (FSH) and

androgens. The androgens are secreted by the Leydig cells,

which are strategically positioned in the testicular

interstitial tissue, between tubules. Since FSH and

testosterone are known to regulate spermatogenesis, the

importance of Sertoli cells in mediating and regulating the

process of sperm production has been emphasized by several

investigators (3-6).

Cyclic Activity

In support of the controlling influence Sertoli cells

impose on spermatogenesis are the variations in Sertoli cell

biochemical activities observed in different stages of

spermatogenesis. There are also cyclic morphological changes

in the shape of the nuclei of the Sertoli cells and in the

abundance and distribution of lipid droplets (7). Recent

morphometric analyses have shown cyclic decreases in Sertoli

cell vessicle volume density and increases in the

endoplasmic reticulum volume density (8). These observations

are indicative of active and variable cellular activities.

The germ cells also exhibit stage-specific cyclic changes

with drastic morphological alterations in size, density,

nuclear arrangement, and acrosome development (9).

The constituent protein changes in spermatogenic cells

reflect the changes found in spermatogenic cell mRNA

populations with differential gene expression during

spermatogenesis (10). For example, the expression of the

cellular protooncogenes N-,Ki-, and Ha-ras has been found to

be differentiation specific during mouse germ cell

development (11,12). Stage specific expression of these

protooncogenes suggest that these genes normally participate

in the differentiative process. N-ras mRNA is specifically

detected in postmeiotic early spermatids, whereas the Ki-ras

transcript is expressed mainly in the meiotic pachytene

spermatocytes. On the other hand, Ha-ras transcript is

reported to be low in both meiotic and postmeiotic cells.

The absence of transcripts of N-ras in mutant sterile mice

(Sl/Sld) which lack germ cells, yet its presence in the

mutant sterile quaking (qk) mice which have spermatids,

supports the idea that these protooncogenes and their

prenylated protein products play a role in germ cell

development. Another example of differential gene expression

during spermatogenesis is a testis-specific transcript of

rat farnesyl diphosphate synthase that was detected by RNA

blot analysis and in situ hybridization (13). This

transcript is exclusively expressed in round spermatids at

stages 7 to 8 in the seminiferous epithelium. Mouse

protamine genes are also solely expressed in haploid

spermatids (14,15).

Extensive chromosome activity occurs during meiosis and

spermiogenesis as described by Monesi (16). RNA synthesis

undergoes a rapid increase in synthesis to a peak in late

pachytene spermatocytes, followed by a decline in diplotene

and during diakinesis. The RNA molecules synthesized in

meiosis, "meiotic RNA" remain associated with the

chromosomes until diakinesis when they are rapidly released

into the cytoplasm. This is the primary RNA present in the

spermatid cytoplasm and its protein products most likely

direct differentiation in spermiogenesis since this time is

also characterized by genetic inactivation and arrest of

nuclear protein synthesis.

The cell types of the seminiferous epithelium exhibit

high metabolic activity with increased protein production

and secretion. Enzyme activities of the Sertoli cells and

the spermatogenic cells have been shown to vary depending on

the stage of the cycle of the seminiferous epithelium. For

example, both acid phosphatase and thiamine pyrophosphatase

activities were found to peak in stages VII and VIII, with

little activity in stages IX through II (17). The Sertoli

cells secrete both testes-specific and serum proteins (18),

many of which have been found to be secreted in a stage-

specific, cyclic manner. For example, androgen binding

protein (ABP) (6), plasminogen activator (19), and

testicular transferring (20) show fluctuations in secretion

rate with specific stages of spermatogenesis. Fucosylation

during the post-translational modification of proteins,

particularly membrane glycoproteins, are continually

changing in the differentiating germ cells (21). Differences

in a host of cell specific proteins in the different stages

and different cell types has been noted. For example, the

somatic cells including Sertoli cells contain lamins A, B,

and C, whereas the spermatogenic cells have lamin B, but do

not reveal immunologically reactive lamins A and C (22).

Enzymes in isoprenoid biosynthesis have also been shown

to change. Potter et al. observed increased rates of [14C]-

acetate incorporation into cholesterol in pachytene stages

of spermatogenesis and high rates of incorporation into

dolichol in differentiating spermatocytes, particularly the

pachytene spermatocytes (23). Experiments performed in this

lab have shown a cyclic change in the specific enzymatic

activity of 2,3-dehydrodolichyl diphosphate synthase in

enriched spermatogenic and Sertoli cell populations (24,

25). The highest activity occurs in 23 day old rats and is

also attributed to the appearance of pachytene

spermatocytes. The fact that spermatogenesis entails

cytochemical changes in gene expression, enzyme activities,

protein secretion, protein distribution and protein

modification is therefore well documented.

Isoprenoid Biosynthesis

The rate-controlling enzyme for isoprenoid biosynthesis

is 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) reductase

(see Fig. 1-1). The product of this enzyme reaction is

mevalonate, the precursor of many isoprenoid groups.

Mevalonate is phosphorylated and decarboxylated to form

isopentenyl diphosphate (IPP), a C5 isoprene intermediate.

Figure 1-1. The Isoprenoid Biosynthetic Pathway

HMG CoA Reductase is the rate-limiting enzyme for the
production of isoprenoids. It is inhibited by the
pharmacological agent mevinolin.

Acetyl CoA



Mevalonic Acid


Diphosphate Dimethylallyl Diphosphate
(C5) (C5)

Geranyl Diphosphate (Clo)

Cholesterol, +-
other sterols

Faresyl Diphosphate (C,,) -+ Farnesylated Proteins

Dolichol <- Geranylgeranyl Diphosphate (C20) -- Geranylgeranylated


Dimethylallyl diphosphate is synthesized by the

isomerization of IPP and then combines with IPP through

three steps, ionization, condensation, and elimination to

form the Ci0 compound geranyl diphosphate. Similarly, IPP

reacts with geranyl diphosphate to give the C15 product,

farnesyl pyrophosphate (FPP), and with FPP to give the C20

product geranylgeranyl diphosphate (GGPP) (26, 27). FPP

occupies a central position in the isoprenoid pathway for

the formation of several end products of mevalonate

metabolism including cholesterol, dolichol, and ubiquinone,

each essential for normal cell function. Cholesterol is a

membrane component, a bile acid precursor, as well as a

steroid hormone precursor. Dolichol is a sugar carrier

important in N-linked glycoprotein biosynthesis and

ubiquinone is a necessary redox catalyst in electron

transport. Other non-sterol mevalonate-derived products

include isopentenyl adenosine, heme a, farnesylated and

geranylgeranylated proteins. So far, the only known function

for all-trans-GGPP in mammalian cells is the modification of

proteins to form geranylgeranylated proteins. The synthesis

of trans,trans,cis-GGPP as an intermediate in dolichol

synthesis is membrane-associated (28, 29).

The synthesis of mevalonate is strictly regulated in

the cell in order to maintain isoprenoid availability

without over-producing sterols (30). HMGCoA synthase, HMGCoA

reductase, farnesyl diphosphate synthase (prenyl

transferse, and squalene synthase are feedback regulated

transcriptionally by cholesterol, the main metabolite of

mevalonate (30-33). Cholesterol also inhibits expression of

the LDL receptor gene in order to limit the contribution of

cholesterol from plasma LDL to the cellular cholesterol.

HMGCoA synthesis is further inhibited by a non-sterol

mevalonate-derived isoprenoid at both translational and

post-translational levels (34-36). The latter is achieved by

increasing reductase degradation. Brown and Goldstein have

speculated that the responsible isoprenoid product is a

prenylated protein (37), although other acute processes may

be involved (38). It is also speculated that the non-sterol

pathways have higher affinities for mevalonate metabolites

(e.g. FPP) than the sterol pathway, which would display yet

another measure of control to ensure synthesis of the non-

sterol isoprenoid products especially when mevalonate is

limiting (30).

Inhibition of HMG CoA reductase and hence the synthesis

of mevalonate prevents protein prenylation and blocks cell

growth (38-44). The drugs compactin and mevinolin are

competitive inhibitors of HMGCoA reductase (45) and have

been shown to halt cell entry into the S phase of the cell

cycle (30, 36, 46). Subsequent addition of mevalonate

restores growth to the drug treated cell cultures. The

addition of serum lipoproteins or cholesterol can reduce the

amount of mevalonate required, but not replace it. This

suggested a requirement for a separate, non-sterol product

of mevalonate for cell cycling (47).

In 1984, Schmidt et al. first showed post-translational

incorporation of isoprenoids into cellular proteins using

radiolabeled mevalonate (48). After mevinolin treatment, the

addition of radiolabelled mevalonate resulted in enhanced

labelling of proteins (49). Sinensky and Logel then

suggested that a halt in DNA replication in mevalonate

starved Mev-1 cells was due to the failure to isoprenylate

proteins (50). Since then, much emphasis has been focused on

identifying the isoprenylated proteins, the nature of

isoprene attachment and the function of the modification.

Protein Isoprenylation

The secreted fungal mating factors were the first

isoprenylated polypeptides identified, and the isoprenoid

moiety was identified as farnesyl (see Fig. 1-1). The first

organism discovered to contain a prenylated peptide was

Rhodosporidium toruloides (51); however, the mechanism of

farnesylation has been studied in particular in the mating

a-factor of Saccharomyces cerevisiae (52). Immediately after

translation, the mating a-factor peptide undergoes several

chemical modifications which are similar to those found in

the p21ras protein processing. These events involve

attachment of a farnesyl group to the thiol of the cysteine,

four amino acid residues from the carboxy terminus.

Subsequent proteolytic cleavage removes the three carboxy

terminal amino acids and leaves the farnesylated cysteine at

the carboxy terminus, which is then methylated. This series

of post-translational modifications is necessary for

secretion of the yeast a-factor (42). The carboxy terminal

sequence termed the CaaX motif (C is cysteine, a is usually

an aliphatic amino acid, and X is variable) has been found

in many other isoprenylated proteins, including the ras

family of guanine nucleotide binding proteins (G-proteins),

that share the same processing events (42). The term CaaX

will continue to be used here; however, it has been reported

that the aliphatic amino acid is not a strict requirement

(especially for the amino acid adjacent to the cysteine)

(53, 54) as was originally suggested. Attachment of the

isoprenoid moiety is required for membrane association and

biological function of the ras proteins (55). Several

isoprenylated GTP-binding proteins, distinct from the ras

family, have been identified in MEL cells by Maltese et al.

(56). The nuclear envelope protein, lamin B in Swiss 3T3

cells, HeLa cells, CHO cells, and MEL cells also contains a

farnesyl group (41, 49). The farnesyl group of prelamin A

has been implicated to function in nuclear localization and

target assembly of this protein to the nuclear envelope

(57). Prelamin A subsequently loses the isoprenoid moiety

concurrently with a 2kD mass reduction (40, 44). Other

proteins have also been ascertained to contain the farnesyl

moiety. These include the y subunit of transducin (58) and

the y subunit of photoreceptor GTP-binding protein (59). It

has been determined that farnesylation of the latter is

necessary for GTP binding. Also, full enzymatic activity of

rhodopsin kinase was found to require farnesylation and c-

carboxyl methylation (60).

It soon became apparent, however, that a majority of

the cellular prenylated proteins were linked with

geranylgeraniol (61, 62). The y subunits of trimeric G

proteins (63-65) and many low molecular weight ras-like GTP-

binding proteins were found to be geranylgeranylated. The

low molecular weight G proteins include G25K (Gp) (66), rac

1, rac 2, and ral A (53), rap 1A (Krev-l) (67), and smg

p25A. The latter was found to be modified with two

geranylgeranyl moieties and a methyl ester (68). Many of the

low molecular weight GTP-binding ras-like proteins that are

geranylgeranylated have carboxy terminal sequences that vary

from the CaaX motif (i.e. XXCC, CCXX, or XCXC). The rab

family of proteins, including rab 1B, rab 2, rab 5, and rab

6 are also geranylgeranylated (69). The rab proteins have

been suggested to participate in the movement of

intracellular vessicles (70). These proteins are clearly

required in nuclear assembly, signal transduction, and

vessicle transport; which are processes necessary for normal

cell function as well as proliferation.

Protein Prenyl Transferases

Three enzymes have been purified that recognize

specific CaaX or other cysteine containing sequences which

are targets for prenylation. The protein farnesyl

transferase (PFT) was purified to homogeneity from rat (71)

or bovine brain (72) as an c/B heterodimer. Each subunit of

the rat brain enzyme have been subsequently cloned by the

Brown and Goldstein group (73, 74). The molecular masses of

the a and B subunits are 47 kDa and 45 kDa, respectively.

Cross-linking experiments implicate the B subunit in

functionally binding the protein substrate (75). PFT

preferentially farnesylates proteins or peptides with the

CaaX sequence where X is Ala, Ser, Gln, Cys, or Met, but

most commonly Ser or Met (54, 72, 76-78). Protein

geranylgeranyl transferase I (PGGT-I) purified from bovine

brain is an o/B heterodimer (79) which geranylgeranylates

proteins with the CaaX sequence where X is Leu (72, 76-78,

80). Selective prenylation of CaaX tetrapeptides indicates

that only the four terminal amino acids of the protein

substrates are necessary determinants for prenylation by PFT

and PGGT-I (78). Antibodies to the a subunit of rat brain

PFT cross react with the a subunit of PGGT-I indicating that

these prenyl transferases share a common subunit (75).

Genetic studies in Saccharomyces cerevisiae also indicate a

common genetic feature between PFT and PGGT-I since

mutations in ram2 give decreased activity for both enzymes

(78, 81, 82). In contrast, mutations in dprl/raml are

specific for PFT, indicating the dprl/raml gene is

responsible for only PFT activity (78, 81). Correlatively,

CDC43/CAL1 was found to be essential for PGGT-I and not PFT

activity (81). Therefore, the yeast protein RAM2 is

homologous to bovine PFT a, whereas RAM1 and CDC43/CAL1

share homology with the B subunit of PFT and PGGT-I,

respectively (83).

The yeast gene BET2 shows homology to RAM1 and CDC43

genes, but bet2 mutants can still prenylate proteins with

the CaaX sequence. On the other hand, bet2 mutants do not

prenylate the GGCC-containing YPT1 protein (83). A third

enzyme, PGGT-II or Rab geranylgeranyl transferase, comprised

in part of the BET2 protein, geranylgeranylates proteins

with carboxy terminal sequences other than CaaX. These

sequences are CXC (such as rab 3A, rab 4, and rab 6), GGCC

(such as rab 1A, rab 1B, and rab 2), and CCXX (such as rab

5) (68, 84-87). At least the rab 3A protein has been shown

to contain two geranylgeranyl moieties (68). However,

peptides of the carboxyl terminus of these proteins do not

compete for protein geranylgeranylation, indicating the

interaction between PGGT-II and its protein substrate is

more complex than that exhibited by PFT and PGGT-I for their

CaaX subsrates (69, 72, 84). PGGT-II activity from rat brain

cytosol separates into two components, A and B (86).

Component B consists of two polypeptides of 60 and 38 kDa

whose amino acid sequences resemble in part those of the a

and B subunits of ras PFT. Component A is a 95 kDa protein

with no counterpart in PFT. Brown and Goldstein have

speculated that there are a family of A components that

recognize the variable protein substrates and are tissue

specific as well (88).


Earlier work showed that enzymes and intermediates of

the isoprenoid biosynthetic pathway changed during

differentiation. For example, the availability of the

isoprenoid dolichyl phosphate has been found to be rate-

limiting for synthesis of glycoproteins and a possible

regulatory component in developmental processes (25, 89-91).

The activity of one of the enzymes required for dolichyl

phosphate synthesis, the prenyl transferase 2,3-

dehydrodolichyl diphosphate synthase, has been shown to

cycle in coordination with increased concentration of

dolichyl phosphate during spermatogenesis (25). Furthermore,

Kabakoff et al. report the level of dolichyl phosphate

available for N-linked protein glycosylation correlates with

cell proliferation during mevalonate depletion and repletion

(92). They suggest the synthesis of dolichyl phosphate may

be rate-limiting for cell proliferation, although they do

not rule out the role of other isoprenoid products in

regulation of cell growth, such as prenylated proteins.

Characterization of the prenyl transferases involved in

isoprenoid attachment to proteins has recently become of

interest since it has been speculated that this post-

translational modification is necessary for cell

proliferation, yet has not been studied in a developmental

system. As described above the expression of the cellular

protooncogenes N-, Ki-, and Ha-ras is stage specific

suggesting that these genes and their prenylated protein

products normally participate in the differentiative process

(11, 12). A recent report, that may have significance in a

developmental system, described cell cycle dependent changes

in prenylated proteins in HepG2 cells (93). Increased

protein prenylation occurred at the Gl/S interface of the

cell cycle, precisely where cell cycling ceased following

mevalonate deprivation. Maltese and Sheridan, however,

suggested that the profile of mevalonate labelled proteins

in a given cell line is not altered by malignant

transformation (49). Since the isoprenoid attachment to the

ras family of proteins is required for ras-mediated

transformation (55) and prenylated proteins pertain to the

non-sterol mevalonate products necessary for cell cycling

(47, 50), one can speculate that protein prenylation is

required in the mitotic and meiotic events during the

differentiative process of spermatogenesis. The recent

developments on protein prenylation prompted me to examine

changes in prenyl transferase activities, modes of protein

prenylation, and the types of proteins prenylated during



The main objectives of this dissertation are to study

the activities of the protein prenyl transferases and the in

vivo labelling of proteins with polyprenols during

spermatogenesis in the rat and correlate the findings with

known cellular transformations.

Specific objectives are the following;

A. Optimize the assay conditions for protein farnesyl

transferase (PFT) and protein geranylgeranyl transferase-I

(PGGT-I) in cytosolic fractions from testicular cells.

B. Determine the specific activities for each enzyme in

the cytosols of the testes of different aged rats.

C. Determine the specific enzymatic activities in the

cytosols of different spermatogenic cell types (particularly

the pachytene spermatocytes and the round spermatids) and in

seminiferous epithelium from rat testes.

D. Optimize labelling of proteins of the seminiferous

epithelium with [3H]-mevalonic acid.

E. Analyze the polyprenols liberated by methyl iodide

treatment of labelled proteins in the seminiferous

epithelium of rats of different ages and in different

spermatogenic cell types.

F. Analyze selected labelled proteins of the

seminiferous epithelium from rats of different ages and from

different spermatogenic cell types by one-dimensional and

two-dimensional gel electrophoresis.

Chapter II of this dissertation describes the studies

toward fulfillment of objectives A through C. Chapter III

describes the studies toward achievement of objectives D

through F.



Differentiation-dependent changes in protein prenyl

transferases have not been described in any system.

Spermatogenesis is a well described process of

differentiation which we have chosen as our model system. In

order to determine the capacity for testicular cells to

prenylate proteins, the enzymatic activities of PFT and

PGGT-I in rat testes were first characterized. These

activities have been localized by others to the cytosol in

testes and other mammalian tissues (41, 43, 48, 71, 76-78,

80, 82, 94-96). The highest levels of PFT activity have been

described in rat and pig brain (94,95) but the levels of PFT

and PGGT-I in testes has not been described. Chen et al.

(93) have reported a large amount of PFT a subunit mRNA in

the rat testes, but only low levels of PFT B subunit mRNA

have been described (74). Age dependent changes in PGGT-I

activity have been reported in rat brain, liver, kidney and

heart where the activity increased from birth and reached a

plateaue at about 20 days of age (96).

The methods for the PFT assay were first described by

Schmidt et al. (48) and Farnsworth et al. (41). The PGGT-I

assays have been described by Seabra et al. (76), Yokoyama

et al. (77), Yoshida et al. (78), and Joly et al. (96).

These procedures have been modified, as described below, to

optimize the analysis of protein prenylation in the soluble

fraction of testicular samples obtained at different stages

of spermatogenesis. This work describes the use of

radiolabelled allylic diphosphates with either unmodified

recombinant protein or biotinylated peptides to detect


This chapter shows 1) optimal parameters for assaying

testicular PFT and PGGT-I; 2) separate activities of these

closely related enzymes in the same cytosolic preparation;

3) age-dependent changes in the levels of PFT activity in

whole, decapsulated testicular cytosols and similar age-

dependent changes in the levels of both PFT and PGGT-I

activity in the cytosols of spermatogenic cells; and 4)

comparisons of PFT and PGGT-I activities in different

spermatogenic cell types. The age-dependent changes for both

activities in spermatogenic cells show a 2-3 fold increase

in specific activity from 9 to 23 days of age and a similar

decrease from days 23 to 35. The evaluation of enzyme

specific activities in different spermatogenic cell

populations of rat testes show similar activities in

pachytene spermatocytes and round spermatids yet higher PFT

activity in a third cell population. Possible roles for

regulation of these enzymes at 23 days and in the third cell

population are discussed.

Materials and Methods


Recombinant p21H-ras (valine 12) was the generous gift of

Burroughs-Wellcome Research Laboratories, Burroughs Wellcome

Co., NC. Recombinant MBP-G23K, MBP-rab 1B, MBP-rab 5, and

MBP-rab 6 were gifts of W.A. Maltese, Weis Research Center,

Geisinger Clinic, Danville, PA. The peptides, biotinyl-Lys-

Thr-Lys-Cys-Val-Ile-Ser (Bt-KTKCVIS) and N-acetyl-KTKCVIS

were synthesized by the Protein Chemistry Core Facility,

Interdisciplinary Center for Biotechnology Research,

University of Florida. Biotinyl-Lys-Lys-Phe-Phe-Cys-Ala-Ile-

Leu (Bt-KKFFCAIL) was generously provided by Dr. A. Joly,

University of California Los Angles. [3H]-Geranylgeranyl

diphosphate (8-15 Ci/mmol) and [3H]-farnesyl diphosphate (20

Ci/mmol) were obtained from Dupont-NEN or American

Radiolabelled Chemicals, Inc. Other reagents and enzymes not

otherwise described were obtained from Sigma Chemical Co.

Animal Groupings

In order to have sufficient data for statistical

analysis in each experiment, at least two rats were used for

each prepuberal age group tested. Studies with younger rats

required more animals. For example, experiments with rats

aged 9-18 days, 19-26 days, and 27-40 days, required 8-12

rats, 5-8 rats, and 2-4 rats, respectively. The excised

testes from each aged group were pooled together for enzyme

assay or subsequent cell separation. The data presented in

each figure are usually means determined by two or three


Isolation of Seminiferous Epithelium. Spermatogenic Cells,
and Spermatozoa

Rats were lightly anesthetized with methoxyflurane,

decapitated with a guillotine, and the testes excised. The

testes were decapsulated by cutting and removing the tunica

albuginea while the tubules were gently expressed. The

spermatic artery was removed and the tissue placed in a 125

ml siliconized flask containing 50 ml of McCoy's medium. The

tubules are washed by unit gravity settling and aspirating

of the supernatant. The seminiferous epithelium was prepared

from the decapsulated tubules by a modification of a two-

step enzymatic method described by Romrell for the mouse

(97). First, 30 ml of 1 mg/ml collagenase and 2 jg/ml

deoxyribonuclease (DNase) in McCoy's medium at 37*C was

added to the tubules; the flask was covered and placed in a

370C water bath for 50 min, with swirling and shaking every

5 min. At the end of the incubation, the suspension was

washed 3 times by unit gravity settling for 2-3 min in 3

changes of 50 ml McCoy's medium. Then 30 ml of 2 mg/ml

trypsin and 2 lg/ml DNase in McCoy's medium was added and

swirled every 5 min in a 37C water bath for 15 min. The

trypsin was immediately neutralized by the addition of equal

units of trypsin inhibitor in 20 ml McCoy's media. The cells

were then mechanically dispersed by pipeting gently 50 times

with a flamed, siliconized pasteur pipet. The spermatogenic

cells were isolated by filtering the cells through a 74g

nylon mesh (Small Parts, Inc.) to remove Sertoli cell clumps

(98). The spermatogenic cells were washed in McCoy's media

three times by centrifuging at 200 x g for 5 min.

Spermatozoa were isolated from the epididymis by cutting the

tissue 8 times followed by suspension in 20 ml of McCoy's

media. The tissue was allowed to settle by unit gravity and

the media containing the mature sperm was removed. All

cellular populations were counted with a hemocytometer and

the viability determined by trypan blue exclusion. The

viability was always 95 to 98%.

Spermatogenic Cell Fractionation

The STA-PUT unit gravity procedure for separating

spermatogenic cells was performed similarly to the process

described by Romrell et al. (98). The whole procedure was

carried out at 4C. A diagram of the procedure is depicted

in Appendix A. The sedimentation chamber was initially

filled with 20 mls of McCoys medium. The cell suspension

containing 108 spermatogenic cells in 40 mis of 0.5% BSA

(Sigma Fraction V, lot# 12H0182) in McCoys medium was

introduced into the chamber at a flow rate of 10 ml/min,

followed by a linear gradient of 2% to 4% BSA in McCoys

medium (2200 ml total volume). Five minutes after loading

the cell suspension, the flow rate was increased to 40

ml/min. Fractions (10 ml) containing the separated

spermatogenic cells were collected at a rate of 10 ml/min

starting 3 h after loading the cell suspension. Within five

hours after introducing the cell suspension to the chamber,

the cell collection was finished. Samples from each fraction

were examined by Nomarski differential interference and

phase contrast microscopy. The fractions consisting of early

spermatids (stages I through VIII) and pachytene

spermatocytes were pooled separately and washed three times

with McCoys medium by centrifuging at 200xg for 10 min.

Subcellular Fractionation and Preparation of Cytosolic

Cytosolic fractions were used as the enzyme source.

Isolated spermatogenic cells or decapsulated testes from

rats of different ages were suspended in 50 mM Tris-HCl, pH

7.5, 10 mM dithiothreitol, 1 mM PMSF, 10% glycerol, then

disrupted in a Dounce homogenizer by twenty strokes with a

type A pestle. Particulate material was removed by

consecutive centrifugations at 9,000 and 100,000 x g,

resulting in the cytosolic fraction. The microsomal fraction

was collected from the pellet of the 100,000 x g

centrifugation. Protein was estimated by the Biorad Protein

assay using bovine serum albumin as a standard.

Protein Prenyl Transferase Assay

In the PFT assay, [3H]-farnesyl diphosphate is

incubated with either recombinant p21ra" or a biotinylated

heptapeptide, Bt-KTKCVIS, whose sequence is similar to the

carboxy terminal sequence of K-ras-2B (38) (the C-terminal

Met was replaced with Ser for easier chemical synthesis).

Similarly the PGGT-I assay was performed with [3H]-

geranylgeranyl diphosphate and recombinant chimeric proteins

or a biotinylated octapeptide Bt-KKFFCAIL, whose sequence

was designed after the carboxy terminal sequence of the y

subunit of bovine brain trimeric G proteins (99, 100). A

high concentration of MgCl2 (5mM) and a low concentration of

ZnClz (25 MM), which had been found by others to be

stimulatory for both enzyme activities, were used in these

experiments. PMSF was routinely added to the assay.

Initially, pepstatin A and leupeptin were added to protect

the protein from endogenous proteases and NaF was added to

protect from endogenous phosphatases. However, there were no

major changes in the extent of formation of radiolabelled

enzymatic products when pepstatin A, leupeptin, and NaF were

omitted from the assay.

The dependence of PFT and PGGT-I activities on

cytosolic protein concentration with appropriate protein and

peptide substrates is depicted in Figure 2-1,A,B,C, and D.

In panel A and B, a linear response of PFT activity was

detected for up to 125 gg protein with p21ras or for up to 25

Ag protein for Bt-KTKCVIS, respectively. PGGT-I (Panel D)

did not show a linear response when less than 50 gg of

protein were used. On the other hand, higher protein

concentrations did not lead to significant increases in

activity. Therefore, 50 pg of protein was selected to assess

PGGT-I activity. In the assays with the Bt-peptide and

allylic substrate concentrations were saturating. This was

not the case with the recombinant p21r" protein. Previous

work with MOLT IV cells suggested that 11.4 pM p21ras is not

saturating for PFT activity (101), which explains at least

in part the lower levels of activity for the protein than

with peptide substrates. Since the source of the recombinant

protein was limiting, the biotinylated peptide was

preferentially used for most of the experiments to follow.

Biotinylated peptides as prenyl acceptors. PFT and

PGGT-I were assayed using biotinylated peptides Bt-KTKCVIS

and BT-KKFFCAIL, respectively by a modification of the

method of Farnesworth et al. (41). Each 25 .l reaction

mixture contained 25-125 gg of cytosolic protein, 60 mM

Tris-HCl, pH 7.5, 25 gM ZnCl2, 5 mM MgC12, 2.0 mM

dithiothreitol, 0.5 mM PMSF, 2% glycerol, 30 gM Bt-KTKCVIS

Figure 2-1. Cytosolic Protein Dependence for PFT and PGGT-I

Incubations contained various concentrations of cytosolic
protein from decapsulated testes of a 70 day old rat (Panels
A, B, and D) or germ cells and seminiferous epithelium from a
23 day old rat (Panel C). PFT activity was assayed as
described in the methods with 11.4 iM recombinant p21ras (Panel
A) or 30 AM Bt-KTKCVIS (Panels B and C). PGGT-I activity was
assayed with 30 jM Bt-KKFFCAIL (Panel D). PFT assays were
performed in duplicate and PGGT-I values were averaged from
triplicate assays.

ug cytosolic protein

or Bt-KKFFCAIL, and 0.4 AM [3H]-t,t-farnesyl diphosphate

(FPP) or 2.0 AM [3H]-all trans-geranylgeranyl diphosphate

(GGPP). Incubations were carried out at 37C for 1 h

followed by the addition of a suspension of avidin-agarose

beads in 0.5 M NaCl, 0.02% NaN3. The tubes were vortexed

repeatedly during a 15 min interval, then the beads were

washed three times with 1 ml of RIPA buffer (10mM Tris-HCl,

pH 7.5, 5 mM EDTA, 1% Triton X-100, 0.1% SDS, 0.1 M NaCI,

0.01% NaN3, 1 mM PMSF, and 1 mM EGTA) and once with 1.0 ml

phosphate buffered saline (PBS). Finally the beads were

suspended in 1.0 ml of PBS and collected on a 25 mm Fisher

Scientific Metrical membrane filter (0.45 mm mesh) with a

Hoefer filtration unit. Each tube was washed three times

with 1 ml of PBS and the washes passed over the filter. The

filter was then solubilized in 5 ml of Ready Protein+

scintillation cocktail and analyzed for radioactivity.

Routine control assay mixtures contained no biotinylated

peptide and were subtracted out for determining activity.

Initially, control assays containing no enzyme were

performed which gave high background levels and were thus


Recombinant protein as prenvl acceptor. PFT was assayed

in some cases using recombinant p21"-ras as prenyl acceptor

(71). Each 25 ~l reaction mixture contained, 25-125 gg of

cytosolic protein, 50 mM Tris-HCl, pH 7.5, 25 AM ZnC12, 100

jM MgCl2, 0.7 mM dithiothreitol, 0.5 mM PMSF, 2% glycerol,

10 mM KC1, 11.4 MM p21H-ras and 0.4 AM [3H]-t,t-farnesyl

diphosphate. After 1 h incubation at 37C, the reaction was

stopped by the addition of 0.5 ml SDS (4%, w/v) followed by

0.5 ml ice-cold 30% (w/v) trichloroacetic acid (TCA) with

mixing by vortex after each addition. The protein was

allowed to precipitate at -20C for 60-90 min and collected

on a 25 mm filter, washed three times with 1.0 ml of 30% TCA

and analyzed for radioactivity as described above. Control

assay mixtures omitted p21H-"s.

PGGT-I and PGGT-II were measured with recombinant

maltose binding protein (MBP) chimeric proteins, MBP-G25K,

MBP-rab 1B, rab 5, and rab 6 according to Kinsella and

Maltese (69). A 50 ~l reaction mixture contained 200 Ag of

cytosolic protein, 25 mM Tris-HCl, pH 7.5, 25 MM ZnC12, 10

mM MgC12, 10 mM dithiothreitol, 50 AM leupeptin, 0.1 AM

pepstatin, 50-170 pmol of recombinant protein, and 2.0 pM

[3H]-t,t,t-geranylgeranyl diphosphate. After 1 h at 37C,

the reaction was stopped by the addition of 500 Al ice cold

acetone and subjected to centrifugation at top speed for 15

min in a microfuge. The pellet was dissolved in Laemmli

sample buffer and subjected to SDS-PAGE (102). Slices (3 mm)

for each lane were incubated with 250 Ml of H20z at 60*C for

18 h, mixed with 10 ml scintillation cocktail (Scintiverse

II, 33% triton X-100) and analyzed for radioactivity.


Testicular protein prenyl transferases. Others have

shown that both enzymatic activities have an absolute

requirement for divalent metal ions (71, 94-96). A limited

number of experiments, not detailed here, showed divalent

cation requirements for PFT and PGGT-I similar to those

described in the literature.

PFT and PGGT-I activity was analyzed in subcellular

fractions of rat testicular cells. The cells were

homogenized in a hypotonic buffer containing 10% glycerin

and then subjected to centrifugation as described in the

methods to yield cytosolic and microsomal fractions. The

activities of both PFT and PGGT-I from decapsulated testes

are exclusively located in the soluble fraction (Fig. 2-2).

The specific activity of PGGT-I measured with the

biotinylated peptide Bt-KKFFCAIL was repeatedly 15-40 fold

less than that of PFT measured with Bt-KTKCVIS. The

dependence of farnesylation on the concentration of

testicular cytosolic protein was measured for both

recombinant p21ras and the biotinylated peptide, Bt-KTKCVIS,

using 0.4 gM [3H]-farnesyl diphosphate (Fig. 2-1A,B).

Similarly, the dependence of geranylgeranylation on

cytosolic protein was determined with Bt-KKFFCAIL using 2.0

kIM [3H]-geranylgeranyl diphosphate (Fig. 2-1D). Most of the

subsequent assays utilized 125 pg and 25 gg of cytosolic

testicular protein for PFT assays with p21ras and Bt-KTKCVIS,

respectively; and 50 gg of protein for PGGT-I assays with

Figure 2-2. Cytosolic Localization for PFT and PGGT-I

Cytosolic (C) and membrane (M) protein were fractionated
from decapsulated testes of 40 day old rats as described in
the methods. The incubations for PFT assays contained 30 AM
Bt-KTKCVIS, 0.4 AM [3H]-FPP, and 25 pg of protein; incubations
for PGGT-I assays contained 30 MM Bt-KKFFCAIL, 2.0 pM [3H]-
GGPP, and 50 Mg of protein. PFT assays were performed in
duplicate. PGGT-I values represent triplicate assays.

10 1.0

S PFT PGGT-I 0.8 o
i 8--
0 o
o 0.6 a0
6 .. c

4 0.4
0 a 4

S2 0.2
0- 0

Bt-KKFFCAIL. The specific activity of PGGT-I was about 40-

fold lower than that of PFT in adult rat testicular cells.

Linearity of PFT activity with cytosolic protein

concentration was also established for preparations of

seminiferous epithelium and spermatogenic cells from 23 day

old rats (Fig. 2-1C). The assay was linear up to 25 Ag of

protein for both cell populations and furthermore; both

these cell populations appeared to have the same levels of

PFT activity.

Optimal substrate concentrations for the biotinylated

peptides and the tritiated allylic substrates were examined.

PFT activity showed optimal activity at 30 pM Bt-KTKCVIS

(Fig. 2-3A). A double reciprocal plot gave an estimated K,

of 14 pM (Fig. 2-3A,insert). PFT assays with varying

concentrations of [3H]-FPP showed that 0.4 AM FPP was

sufficient to saturate the enzyme (Fig. 2-3B). The Km for

FPP was established to be 0.2 IM (Fig. 2-3B,insert). Optimal

PGGT-I activity was achieved at 30 iM Bt-KKFFCAIL and 2.0 AM

with GGPP (Fig. 2-4A,B). Because of the lower activity with

PGGT-I and increased scatter in the data, only Km estimates

of 18 iM and 1.7 pM can be made for the peptide and GGPP,

respectively. A comparison of these Kms with those reported

for similar enzymes are depicted in Table 2-1. All assays

were performed at saturating concentrations of biotinylated

peptides and allylic substrates and within the linear

response range for cytosolic protein, except for the

Figure 2-3. Peptide and Allylic Substrate Concentration
Dependence for PFT Activity.

PFT incubations contained 25 pg of cytosolic protein from
decapsulated testes of a 26 day old and were assayed as
described in the methods with various peptide or allylic
concentrations. The concentration of Bt-KTKCVIS (Panel A) or
[3H]-FPP (Panel B) was varied as indicated. Insets in Panel A
and Panel B show double reciprocal plots for the PFT
substrates. All points represent the average of duplicate
assays. Error bars not seen lie within the symbol.



V0 I I I I
0 10 20 30 40 50


0 0.2 0.4 0.6 0.8 1.0

pM [3H]-FPP


-0. 0.2 0.6 1.0






Figure 2-4. Peptide and Allylic Substrate Concentration
Dependence for PGGT-I Activity.

PGGT-I incubations contained 50 ug of cytosolic protein
from decapsulated testes of a 26 day old rat and were assayed
as described in the methods with various concentrations of
peptide or allylic substrates. The concentration of Bt-
KKFFCAIL (Panel A) or [3H]-GGPP (Panel B) was varied as
indicated. All data points represent the average of triplicate


0.32- O

0.16-- T







15 30 45



0.46 I
I 0

0.23T 0I

Ar. *

0 1.0 2.0 3.0

pM [3H]-GGPP



A% I I






Table 2-1

Approximate Ks (uM) of Protein. Peptide and Allylic Substrates
for PFT. PGGT-I and PGGT-II'



Protein or Peptide







cytosol, rat testes
cytosol, TT cells
ppC, rat brain
pp, bovine brain
pp, bovine brain
pp, bovine brain
pp, bovine brain
pd, bovine brain
p, rat brain

PGGT-I Protein or Peptide GGPP Source Ref.

Bt-KKFFCAIL 18 1.7 cytosol, rat testes this study
Ras-CAIL 0.5 0.09 pp, bovine brain 72
Bt-KKFFCAIL 5 pp, bovine brain 77
RhoA-CLVL 1 3 pp, bovine brain 78
Ras-CVLL 1.5 0.3 pp, bovine brain 80

PGGT-II Protein





pp, bovine brain
pp, bovine brain
p, rat brain

Incubation conditions for PFT and PGGT-I referenced to this study were the same as
described in the methods.
bRef refers to the reference where the K. information was obtained.
"pp = partially purified
"p = purified to homogeneity


this study



experiment where recombinant protein was used as the prenyl


Time dependence for PFT and PGGT-I activities was

measured with the biotinylated peptide substrates. PFT

activity was linear for at least 60 min (Fig. 2-5). Sixty

min was also an appropriate incubation time for PGGT-I

activity (data not shown).

Evaluation of the ability of related but different

peptides to act as substrates or inhibitors of enzyme

activity is useful in establishing enzyme specificity. Since

PFT and PGGT-I were assayed from the same cytosolic

preparation it was necessary to demonstrate separate

activities. Farnesyl transferase activity showed a strong

specificity for the peptide substrate Bt-KTKCVIS, whereas

Bt-KKFFCAIL, the PGGT-I substrate, was not farnesylated

(Fig. 2-6). In contrast, for assessing geranylgeranyl

transferase activity, Bt-KKFFCAIL was the best substrate but

Bt-KTKCVIS was also geranylgeranylated to the extent of 50%

of Bt-KKFFCAIL. Although cross-reactivity of substrates was

observed, it was concluded that distinct activities of PFT

and PGGT-I are being measured when Bt-KTKCVIS and Bt-

KKFFCAIL, respectively are used as substrates. This

conclusion is discussed in more detail in the discussion

section of the Chapter.

Figure 2-7A shows inhibition of PFT by increasing

concentrations of the acetylated peptide, Ac-KTKCVIS. Only

Figure 2-5. Time Dependence for PFT Activities.

PFT activity was assayed as described in the methods with
30 AM Bt-KTKCVIS, 0.4 AM [3H]-FPP and 125 Ag of decapsulated
testicular cytosolic protein from a 58 day old rat at the time
points indicated. Each time point was performed in duplicate.




0 15 30 45 60

Figure 2-6. Peptide Specificity of PFT and PGGT-I.

PFT and PGGT-I activities in decapsulated testes from 17
day old rats were each assessed using 30 pM Bt-KTKCVIS or 30
jM Bt-KKFFCAIL with [3H]-FPP (and 25 Ag cytosolic protein) or
[3H]-GGPP (and 50 Ag cytosolic protein) as described in the
methods. Controls for PFT and PGGT-I assays in which Bt-
peptide were omitted gave values of 1.7 pmoles/mg and 0.5
pmoles/mg, respectively which have been subtracted. PFT assays
were performed in duplicate and PGGT-I assays were performed
in triplicate.

PFT Activity
with 3H-FPP

PGGT-I Activity

with 3H-GGPP

___I_____..... ..-. --.. .- a-


m :13; n en
6 u.yp
5i u. u

-1.2 (i

-0.9 0.

-0.6 3




3.0 +




Figure 2-7. Inhibition of Farnesylation of Bt-KTKCVIS with N-

PFT activity was assayed with 25 gg of decapsulated
testicular cytosolic protein from a 45 day old rat as
described in the methods with (A) 30 jM Bt-KTKCVIS and varying
concentrations of N-acetyl-KTKCVIS or (B) 30 AM BT-KTKCVIS and
either 120 AM N-acetyl-KTKCVIS or 120 AM N-acetyl-KKFFCAIL.
All assays were performed in duplicate, error bars not seen
lie within the symbol.





0 30 60 90 120


None 120 uM

120 uM

Peptide Inhibitors


15 MM was necessary to inhibit the prenylation of Bt-KTKCVIS

by half, which suggests the acetylated peptide may be a

better substrate for PFT than the biotinylated substrate.

Similar K s indicate that the biotinyl group has little

effect on the suitability of Bt-KTKCVIS as a substrate. The

specificity of inhibition of PFT by peptides is further

demonstrated in Fig. 2-7B, since the same concentration of

Ac-KKFFCAIL failed to appreciably inhibit the prenylation of

Bt-KTKCVIS under conditions where Ac-KTKCVIS gave nearly

complete inhibition.

Age-dependent activity of protein prenyl transferases.

In order to determine if protein prenyl transferase

activities change during distinct stages of spermatogenesis,

prepuberal animals were utilized. This is the only period

when specific cell types emerge for the first time during

differentiation. Cytosolic fractions of the decapsulated

testes of different aged rats were assayed for PFT. PFT

specific activity measured with either recombinant p21H-ras or

Bt-KTKCVIS as polypeptide substrate showed a peak in

activity at 26 days of age when measured with 125 Mg of

protein (Fig. 2-8). This indicates the biotinylated peptide

is a representative substrate for PFT and can be used to

accurately measure changes in the enzyme capacity to

farnesylate. Activity at the peak was 2-3 fold higher than

that observed at earlier or later ages. The specific

Figure 2-8. Prepuberal Age Study of PFT Activity Comparing Bt-
KTKCVIS Peptide with p21H-ras Protein as Substrates.

Activity was assayed using 125 pg of decapsulated
testicular cytosolic protein taken from different aged
prepuberal rats and either 30 pM Bt-KTKCVIS or 11.4 iM p21-ras
as described in the methods. All values for data points
measured with p21H-ras were performed in duplicate; error bars
not seen lie within the symbol. Data points measured with Bt-
KTKCVIS were averaged from 1-3 experiments, shown in Fig. 2-9,
each performed in duplicate.

I p

A-a p2 ras


I |

10 15 20 25 30 35 40


9.0 U

6.0 W



Days of Age




activities of PFT in the cytosolic fraction from animals

aged 45-105 days was not greater than that demonstrated at

17 days of age (Fig. 2-9). It was of interest to determine

if the peak of activity in 26 day old animals was

contributed to by the changing population of germ cells.

Germ cells were isolated from Sertoli cells and interstitial

cells and tested for prenyl transferase activity. An age

dependent peak of PFT activity at 23 days of age was

observed with isolated spermatogenic cells (Fig. 2-10A).

PGGT-I specific activity showed a peak that was coincident

with age to that found for PFT in spermatogenic cells (Fig.

2-10B). Although the specific activity of PGGT-I was twenty-

five fold less than that for PFT, the peak in activity at 23

days of age was still 2-3 fold higher than that seen in

animals several days younger or older.

The other known enzyme that geranylgeranylates

proteins, PGGT-II, was assayed in 17, 23, and 60 day old

animals using chimeras of recombinant maltose binding

protein (MBP) and rab 1B, rab 5, and rab 6, which have the

carboxy terminal sequences indicated in Table 2-2.

Recombinant proteins specifically prenylated by either PGGT-

I or PFT were utilized as controls. These proteins were a

MBP-G25K chimera with a carboxy terminal sequence CVLL and

p21ra" (not a chimera) with the carboxy terminal sequence

CVIS. p21ra" was clearly labelled with [3H]-FPP by cytosol

from 60 day old animals. MBP-G25K was labelled with [3H]-

Figure 2-9. Prepuberal Changes in PFT Activity in Comparison
with Adult Activity.

PFT activity, assayed from decapsulated testes of adult
rats using 125 jg of cytosolic protein and 30 pM Bt-KTKCVIS,
was plotted with data from Fig. 2-8. Numbers in paranthesis
represent the number of assays performed at each age and the
error bars demonstrate variances between experiments. Each
assay was performed in duplicate.

.r 9.0-
c /
o (1)
o' 6.0
( (1)


/1 (2) 1(2)
o (1)
0 I i I6
5 25 45 65


I i



Days of Age

Figure 2-10. Age Dependent Changes in Protein PFT and PGGT-I
Activities in Spermatogenic Cells.

Spermatogenic cells were isolated as described in the
methods. Activity was assayed using 60-80 pg of cytosolic
protein and either 30 pM Bt-KTKCVIS (PFT, Panel A) or 30 AM
Bt-KKFFCAIL (PGGT-I, Panel B). PFT assays were performed in
duplicate; error bars not seen lie within the symbol. PGGT-I
assays were performed in triplicate.







0.08 +




Days of Age






Table 2-2



p21" (-CVIS)


MBP-rablB (-GGCC)

MBP-rab5 (-CCSN)

MBP-rab6 (-GCSC)

Study of Prenylation of Recombinant Proteins'

17 day old 23 day old
(pmoles/mg) (pmoles/mg)

0.320b 1.322b









60 day old







'Incubation conditions for prenylation assays were performed as described in the methods
for recombinant proteins.
bThese data were obtained from PFT assays using TCA precipitation and filter binding.
All others were obtained from radioactivity analysis of dissolved PAGE slices.

GGPP by cytosol from 23 day old animals to a level 1.5- to

2-fold higher than that seen with cytosol from 17 and 60 day

old animals. This substantiates the peak observed for PGGT-I

activity in 23 day old animals measured with the

biotinylated peptide Bt-KKFFCAIL. In contrast, activity for

the PGGT-II enzyme, measured with MBP-rab proteins, were

very low with no age dependent differences. Since the

recombinant proteins were present in E. coli. extracts,

accurate measurements of the protein concentrations were not

possible. Adequate levels of protein substrate were believed

to be present since no effect on the levels of activity were

seen on doubling the amount of E. coli extracts.

The results of these studies were attained at

relatively high concentrations of cytosolic protein (60-125

Mg) which is near the limit of the linear response in some

cases. This could have affected the position and extent of

the peak of activity. However, the peak of PFT activity in

decapsulated testes at 23 days of age was confirmed by

testing at lower protein concentrations (25 Mg). A two-fold

increase in cytosolic activity was observed going from 17

days of age (3.4 0.6 pmoles/mg protein x h) to 23 days

days of age (6.3 0.2 pmoles/mg protein x h) with Bt-

KTKCVIS as substrate. A similar decline was seen going from

23 to 26 days of age (4.2 0.4 pmoles/mg protein x h).

PFT activity differences were also assessed between

decapsulated testes and germ cells at the lower protein

concentration (25 gg). The specific activity of cytosolic

PFT in decapsulated testes from 23 day old rats was 6.3

0.2 pmoles/ mg protein x h, whereas the activity of

spermatogenic cells was 2-fold higher with 13.7 0.9 pmoles

product/mg protein x h. At other ages tested, PFT activity

in spermatogenic cells was also higher than in decapsulated


The contribution of the somatic Sertoli cell to the

age-dependent changes in PFT activity were assessed by

assaying PFT in germ cells and seminiferous epithelium at 8

and 23 days of age (Fig. 2-11). Similar to previous results,

a 2-fold increase in activity occurred from 8 to 23 days of

age in germ cells and seminiferous epithelium as shown in

Fig. 2-1C. PFT specific activities between germ cells and

seminiferous epithelium from 23 day old animals were equal.

Furthermore, since Fig. 2-11 shows that both cell

populations increase in PFT activity with little difference

in specific activity between them at each age tested, this

suggests that the Sertoli cells, which are the somatic cells

of the seminiferous epithelium, have a similar specific

activity and that they cycle in farnesylation capacity

coincidently with the germ cells. Thus, somatic cells make

an equal contribution to activity.

Spermatocenic cell type differences in protein prenyl

transferase activities. In order to assess the spermatogenic

Figure 2-11. PFT Activity in Spermatogenic Cells and
Seminiferous Epithelium.

Seminiferous epithelium and spermatogenic cells were
isolated from the same group of rats as described in the
methods. PFT activity was assayed with 25 pg of cytosolic
protein and 30 pM Bt-KTKCVIS. Bar values represent the average
of duplicate assays.

8 days 23 days

Days of Age










cell types present at various days of age in the prepuberal

rat, data from Zhengwei et al. (103) on quantitation of cell

types in the developing rat testes were converted to give

the percentage of each spermatogenic cell type present

relative to the whole population of spermatogenic cells at

particular days of age. These percentages were plotted (Fig.

2-12) along with PFT specific activities from rats of the

same days of age (data from Fig. 2-10A). At 9 days of age

the spermatogonia are the only spermatogenic cell type

present, so they represent 100% of the germ cells. As the

primary spermatocytes appear at day 15 and rise in numbers,

the spermatogonia decrease relative to the spermatocytes.

The primary spermatocytes peak at 23-25 days as the most

prevalent cell types present. The round spermatids appear at

day 25 and quickly accumulate in numbers. By day 40, just

before puberty, the spermatids are the most highly

represented of the spermatogenic cells. The round spermatids

continue to increase in numbers well past puberty in the

adult rat. These data show a correlation of the rise in PFT

activity with the rise in numbers of spermatocytes.

In order to directly assess the PFT activity in

selected types of germ cells, it was necessary to physically

separate them. The Sta Put unit gravity cell separator

allowed the separation and identification of the pachytene

spermatocytes and the round spermatids. Spermatozoa, mature

sperm, were isolated from the epididymis of the adult rat.

Figure 2-12. Age Dependent Changes of Spermatogenic Cell
Population and PFT Activity.

PFT activity in prepuberal spermatogenic cells from Fig.
2-10, Panel A, is plotted along with the percentage of
indicated cell types present in the spermatogenic cell
population from Zhengwei et al., 1990 (103).


100 al ol germ cells -7.0
S-* spermotogoniao
Sa ~ A spermatocytes -
| 80 s A-A spermatids *5.6 3

4 0.60 -4.2 1

S40- -2.8

S 20- A 1.4
0 -0
5 10 15 20 25 30 35 40 45

Days of Age

Measurements of specific activity of PFT were similar

between adult pachytene spermatocytes and round spermatids

(Fig. 2-13, P and R), whereas the specific activity in

spermatozoa was negligible (Fig. 2-13, SZ). Pachytene

spermatocytes were also isolated from 23 day old animals and

analyzed for PFT activity. The prepuberal spermatocytes

showed a 2 fold higher level of PFT than the adult

spermatocytes, yet the whole spermatogenic cell mix from the

same 23 day old animals revealed four fold higher PFT

specific activity (Fig. 2-13, 23P and 23W). Thus, the

pachytene spermatocytes alone were not responsible for the

increase in PFT specific activity at 23 days. The specific

activities of PGGT-I is not significantly different among

any of these cell types (Fig. 2-14). In order to further

identify the cell type responsible for the peak of PFT

activity in 23 day old spermatogenic cells, pachytene

spermatocytes and all Sta Put fractions eluted before and

after the pachytene spermatocytes were collected, pooled,

and assayed for PFT activity. Figure 2-15 shows that the

cells with the highest activity appear in the pooled

fraction eluted just after the pachytene spermatocytes. This

pooled fraction represents a mixture of cell types all of

which are slightly lighter in density and smaller in size

(as seen by light microscopy) than the pachytene


Figure 2-13. PFT Activity in Isolated Pachytene Spermatocytes
and Round Spermatids in Adult and 23 Day Old Rats.

Pachytene spermatocytes (P), round spermatids (R) and
spermatozoa (SZ) were isolated from 78 day old rats, as
described in the methods. Pachytene spermatocytes (23P) were
isolated from spermatogenic cells of 23 day old rats (23W) by
the same procedure. PFT activity was assayed using 25 pg of
cytosolic protein and 30 gM Bt-KTKCVIS. The data represent
duplicate assays in one experiment, which is representative of
3 separate experiments.





P -Pachytene Spermatocytes
R -Round Spermatids
23P -23 day old Pachytene S.
23W -23 day old Whole
Spermatogenic Cells
SZ Spermatozoa

7_I z_

R 23P


Figure 2-14. PGGT-I Activity in Isolated Pachytene
Spermatocytes and Round Spermatids in Adult and 23 Day Old

Spermatogenic cell types were isolated from 78 and 23 day
old rats, as described in the legend of Fig. 2-13. PGGT-I
activity was assayed using 50 pg of cytosolic protein and 30
MM Bt-KKFFCAIL. The data represent triplicate assays in one
experiment, which is representative of 3 separate experiments.

I I _____________ I














Figure 2-15. PFT Activity Assayed in Pooled Sta Put Fractions
of Spermatogenic Cells from 23 Day Old Rats.

All fractions were pooled as indicated and assayed for PFT
activity with 25 pg of cytosolic protein and 30 pM Bt-KTKCVIS.
Each bar value represents duplicate assays. The data are
representative of 3 experiments.

BP=before PS
P=Pachytene Spermatocyes (PS)
AP-1=1 st pooled
fraction after PS
AP-2=2nd pooled
fraction after PS
W-whole cell mix



P AP-1 AP-2














Testicular protein prenvl transferases. The testicular

PFT and PGGT-I activities are localized to the cytosolic

fraction (Fig. 2-2) as has been described for each of the

protein prenyl transferases described thus far in mammals

and in yeast (41, 43, 48, 71, 76-78, 80, 82, 94-96). The

prenyltransferase with the highest specific activity in

cytosols was PFT, with PGGT-I 15-40 fold less active.

Whereas previous reports are semi-quantitative in nature,

these studies give more quantitative analysis of comparisons

of the levels of PFT and PGGT-I activities. Studies with

partially purified enzymes have shown 8-fold higher levels

of PFT activity with recombinant p2lc-Ha-ras than seen with

PGGT-I activity, measured with p21rho (78). Yokoyama et al.

have also observed 10-fold higher levels of activity in

partially purified PFT than in partially purified PGGT-I

activity measured with biotinylated peptides (77). These

differences in the levels of activity of PFT and PGGT-I are

consistent with reports describing efforts to use peptide-

affinity column chromatography for purification of prenyl

transferases. The columns were a convenient way to

extensively purify PFT (71) but were ineffective for

purifying PGGT-I (77). It was determined that the affinity


of PGGT-I for its peptide substrates was less than that for

PFT (71, 77).

Estimated Ks of PFT for its substrates were determined

to be 14 MM for Bt-KTKCVIS and 0.2 MM for [3H]-FPP, which

are not significantly different from that seen in the

literature (Table 2-1). The Kms of PGGT-I for Bt-KKFFCAIL

and [3H]-GGPP were 18 AM and 1.7 MM, respectively, also

similar to that reported in the literature (Table 2-1).

Separate activities for PFT and PGGT-I were established

by the exhibited differences in substrate specificity (Fig.

2-7). The strong preference of the testicular PFT to

farnesylate CaaX peptides ending in serine and not leucine

is consistent with many prior studies (54, 72, 77, 78). The

observed geranylgeranylation of Bt-KKFFCAIL almost certainly

is a measure of PGGT-I activity since PFT, which

farnesylated this peptide so poorly, is not likely to use

geranylgeranyl diphosphate as an effective donor. Assignment

of the prenyl transferase responsible for

geranylgeranylation of the PFT substrate is more difficult

to assess. Yokoyama et al. (77) showed geranylgeranylation

of a PFT substrate with partially purified PFT and PGGT-I

(PGT). Biotinylated -NPFREKKCAIS (a PFT substrate) was

geranylgeranylated by PGGT-I to the extent of 50% of that of

Bt-NPFREKKCAIL (a PGGT-I substrate). They also showed that

PFT geranylgeranylated Bt-NPFREKKCAIS to some extent (6-fold

less than the PGGT-I), whereas this enzyme modified Bt-

NPFREKKCAIL with geranylgeranyl to negligible amounts.

Therefore, although PGGT-I alone is undoubtedly responsible

for the geranylgeranylation of the Bt-KKFFCAIL peptide, PFT

may be responsible for some geranylgeranylation of the Bt-

KTKCVIS peptide. Several investigators have reported

sequences that are farnesylated by PFT in vitro, but are

found to be geranylgeranylated in vivo (72, 77). Clarke

proposed that this cross-reactivity suggests the relative

concentrations of acceptor proteins and the two allylic

donors may help direct isoprenylation in intact cells (104).

The specificity of PFT was established by comparing the

ability of peptides with different CaaX sequences to serve

as substrates or inhibitors. The acetylated form of the

peptide acceptor (Ac-KTKCVIS) inhibited and possibly

competed with Bt-KTKCVIS for farnesylation in a

concentration dependent manner. Fifty percent inhibition

occurred with 15 gM Ac-KTKCVIS, which suggested the

acetylated peptide bound better than the biotinylated

substrate to PFT. Further demonstration of specificity was

shown by the lack of inhibition by N-Ac-KKFFCAIL, the

acetylated peptide corresponding to the carboxy terminal

sequence specific for geranylgeranylation whereas the N-Ac-

KTKCVIS peptide inhibited essentially 100%.

Age-dependent activity of protein prenyl transferases.

The results show that the specific activities of the

testicular protein prenyl transferases PFT and PGGT-I in the

prepuberal animal vary with animal age. It is probable then

that isoprenylated proteins and protein prenyl transferases

are expressed in a differentiation dependent manner and

possibly a cell specific manner. The activity of another

testicular prenyl transferase, dehydrodolichyl diphosphate

synthase, also showed dependence on animal age (24).

Synthase activity increased in early stages of

differentiation during the spermatogenic process, peaking at

23 days, as shown here for the protein prenyl transferases.

It was concluded that this enzyme was responsible for the

increase in dolichyl phosphate measured during this period

of development. Among the spermatogenic cells, the pachytene

spermatocytes had the highest levels of synthase activity


The advantage of assaying the cytosols from rats of

varying prepuberal ages is that at different times during

the first.wave of germ cell differentiation, distinct stages

may be examined where the spermatogenic cell types present

are well-known and are unaccompanied by more differentiated

cell types. A 2-3 fold peak of PFT activity occurred at 26

days, when measured with either p21as or Bt-KTKCVIS (Fig. 2-

8). The activity decreased similarly from 26 days to 35-50

days of age, a time when the rat reaches puberty, and

remains at low levels in the adult rat (Fig. 2-9).

It was of interest to assess the cellular origin,

somatic or germ, of the peak of activity in 26 day old


animals. This was tested by determining if isolated germ

cells had prenyl transferase activity. A 2-3 fold age

dependent peak of PFT activity was observed with isolated

spermatogenic cells at 23 days of age (Fig. 2-10A).

Interestingly, cytosolic activity from the spermatogenic

cells of 23 day old animals was 2 fold higher than activity

in cytosols from decapsulated testes isolated from the same

pool of animals. This may represent a loss of connective

tissue and accompanying extracellular protein when purifying

the germ cells. Yet the changes in PFT activity in isolated

spermatogenic cells suggests that farnesylation in germ

cells is important for development at this stage of


The age dependent studies performed with PGGT-I

substrates and cytosol from spermatogenic cells also showed

a peak of activity at 23 days of age, which was 2-3 fold

higher than that seen in 17 or 28 day old animals (Fig. 2-

10B). However, the specific activity of PGGT-I at each age

is approximately 25 fold less than that observed for the PFT

enzyme. It appears that the geranylgeranylation of proteins

with the CaaL carboxy terminus is also important in the

developing germ cell at this stage of spermatogenesis.

However, the low activity of PGGT-I suggested that the

enzyme known to geranylgeranylate proteins with alternative

cysteine containing carboxy terminal sequences might be

responsible for protein geranylgeranylation.


Geranylgeranylation of proteins with non-CaaX sequences was

examined with MBP-rab proteins rab 1B (-GGCC), rab 5 (-

CCSN), and rab 6 (-GCSC) at these same stages of

spermatogenesis. The specific activity of this enzyme was

extremely low in all ages tested, 5-10 fold lower than the

PGGT-I activity measured with the recombinant G25K or with

Bt-KKFFCAIL. There were no significant differences in

activity between ages 17 and 23 days of age, but somewhat

lower activity in 60 day old animals. The differences among

the levels of PFT, PGGT-I, and PGGT-II specific activity was

perplexing since most of mammalian prenylated proteins are

geranylgeranylated (61, 62). This apparent contradiction may

be partially explained by the results presented in Chapter

III. The levels of protein prenylation in vivo may be

determined primarily by the amount and type of substrate

proteins available, thus portraying the geranylgeranylating

enzymes as having a high capacity, yet low affinity for the


To further examine the question of somatic or germ cell

origin for the peak of prenyl transferase activity at 23

days, the Sertoli cells were assayed indirectly for age-

dependent PFT activity. The Sertoli cells are indispensably

involved in germ cell development and are known to exhibit

cyclic activity of other enzymes (6, 8, 10, 17, 19-21, 24,

25). The Sertoli cells are difficult to isolate, due to

their tight junctions and their envelopment of germ cells,


and require days of culturing to release the germ cells. The

lengthy culturing period and resulting uncertainty of

subsequent changes in activities coupled with the low yield

of cellular protein from primary cultures rendered this

experiment difficult to conduct and with the possibility of

obtaining questionable results. Therefore, PFT activity in

cytosols from the seminiferous epithelium (Sertoli cells and

germ cells) and germ cells of different aged animals were

compared (Fig. 2-11). A 2-fold rise in activity occurred in

both cell populations from 8 to 23 days of age, with little

decrease from 23 to 35 days. This suggests that both Sertoli

cells and germ cells exhibit increased activity between 8

and 23 days of age. Sertoli cells represent 68% of the cells

in 8-9 day old rat seminiferous epithelium, 58% of those at

23 days, and 38% of those at 35 days of age. Thus, Sertoli

cells contribute decreasing percentage of total protein in

the seminiferous epithelium during testicular maturation

from 8 to 23 days of age, yet the PFT activity of the

seminiferous epithelium matches the increase seen with germ

cells alone. Since there was no difference in the specific

activity between the seminiferous epithelium and the germ

cells at any age, it is most likely that Sertoli cells have

PFT specific activity similar to the germ cells and also

cycle in the production of prenylated proteins in


Spermatogenic cell type differences in protein prenyl

transferases. The work of Zhengwei et al. (103) quantitated

the numbers of spermatogenic cells present at various days

of age in the developing rat. The cell types discussed were

the spermatogonia, the cell types that develop from the stem

cell and that undergo mitotic divisions; the spermatocytes,

the cell types that develop just after DNA synthesis in

prophase of meiosis; and finally the haploid spermatids, the

product of meiosis which undergo extensive morphological

changes as they elongate into the mature sperm. At 9 days,

when PFT activity is low, the spermatogonia are the only

spermatogenic cell type present. The spermatocytes appear

around day 15 and by day 23, when the PFT activity peaks,

they are the most prevalent cell type present. The activity

declines with the presence of the spermatids at day 25 and

levels off in adult animals as the spermatids continue to

accumulate in numbers. The correlation of the appearance of

the spermatocytes and the rise in PFT activity indicated

these cell types to be responsible for the peak in activity.

Therefore, PFT activity was examined in separated pachytene

spermatocytes and round spermatids. Surprisingly, the levels

of activity were the same in both cell types (Fig. 2-13).

This observation seems to be inconsistent with either the

hypothesis that 1) the increase in activity at 23 days is

due to the increase in pachytene spermatocytes or 2) the

subsequent decline in activity is due to the appearance of

more mature, less active spermatids. To further test the

first hypothesis, pachytene spermatocytes were isolated from

23 day old rats and analyzed for PFT activity (of course

round spermatids are not yet present in 23 day old animals).

These cell types had 2-fold higher activity than the adult

pachytene spermatocytes. This is not too surprising since

studies in prepuberal animals are typically more active than

the adult. For example, Potter et al. (23) showed higher

incorporation of [14C]-acetate into dolichol and cholesterol

in prepuberal pachytene spermatocytes than in the adult

pachytene spermatocytes, as well as greater HMG CoA

reductase specific activity. However, the whole

spermatogenic cell population from the same 23 day old

animals demonstrated even higher PFT activity (4 fold) than

the adult pachytene spermatocytes. Therefore the peak in

activity observed at 23 days of age can not be correlated

with the pachytene spermatocytes alone. Apparently, either a

cell type with high activity was eliminated in the Sta Put

separation and selection procedure, or the cells are more

active when present with other spermatogenic cell types. The

latter seems unlikely since there is no evidence of

interaction between spermatogenic cells in vivo, without the

presence of Sertoli cells. The presence of a soluble

stimulatory "factor" can not be ruled out, but also seems

unlikely since the whole spermatogenic cell mixture, which

was not subjected to the Sta Put but was subjected to the

_ __

same washes in fresh medium and is maintained at 4*C,

retains high activity.

To examine whether another cell type present at 23 days

of age gives rise to the increase in PFT activity, all Sta

Put fractions of 23 day old spermatogenic cells were pooled

and assayed for PFT activity (Fig. 2-15). The pooled

fractions with the highest PFT activity contain a mixture of

cell types that are slightly smaller than the pachytene

spermatocytes. The cell type in this mixture responsible for

the high activity is probably not spermatogonia, since these

are a greatly reduced fraction of the germ cells at this

time and the activity is low in 9 day old animals where the

spermatogonia are the only spermatogenic cell type present.

Early spermatocytes, preleptotene, leptotene, and zygotene

spermatocytes may be present in this mixture, yet their

first appearances at days 10-15 is represented with low

levels of activity. The only possible cell type that would

be smaller than the pachytene spermatocytes and present in

23 day old rats is the secondary spermatocyte, the cell type

that has undergone the first division of meiosis and is

preparing for the second division. This is a very short-

lived cell type and remains undetected by the available

means of cell identification. This cell type may also be

particularly sensitive to the "irregular journey" through

the Sta Put, which would explain the high activity in the

whole cell population which was not exposed to the Sta Put.


It seems most probable that the peak in activity at 23 days

of age and the presence of higher activity in a pooled

fraction of the Sta Put is due to the increased activity of

of spermatocytes. The rise of PFT activity from 9 to 23 day

old rat testes was not due solely to the rise in numbers of

the pachytene spermatocytes as noted for other activities

that increase during this time (23, 25). Although the cell

type responsible for the highest activity can not be

unambiguously identified, it may be the secondary

spermatocyte. Undoubtedly, the peak in PFT activity occurs

just prior to, during, or just subsequent to the events of

the meiotic divisions in the development of the rat sperm,

implicating the involvement of prenylated proteins at this

point in spermatogenesis. PFT activity in spermatozoa is

very low. This is not surprising, since development of the

mature sperm involves stripping the cell of most of its


The same cell types (adult pachytene spermatocytes,

round spermatids, 23 day old pachytene spermatocytes, whole

germ cell population from 23 day old animals, and

spermatozoa) were found to have no significant differences

in PGGT-I activity. This may be due to variable

determinations for PGGT-I activity with extremely low levels

of enzyme activity, which are accompanied by high

backgrounds that might obliterate any potential differences.

Therefore, the information gathered on the activity

differences in the isolated cell types for PGGT-I is

somewhat limited.

Less can be said about prenyl transferase activities in

the somatic cell, the Sertoli cell. Sertoli cells respond in

a biochemically different manner to the spermatogenic cells

with which they are in contact (105), so the observed

increase in Sertoli cell PFT activity at 23 days of age may

represent a response to the appearance of primary

spermatocytes. Further work in this area will require the

development of better methods for separating and culturing

larger numbers of Sertoli cells.

An interesting correlation to this study is the work by

Monesi (16) which showed that genetic inactivation and

arrest of transcription was apparent in the spermatids. The

accumulated cytoplasmic RNA was termed meiotic RNA, since it

was transcribed in meiosis and released into the cytoplasm

at diakinesis. The meiotic RNA is particularly long-lived

and sustains the events of spermiogenesis. It is possible

that the PFT and PGGT-I enzymes are made in spermatocytes,

are long-lived, and thus carry the potential for prenylation

activity in to the spermatids, regardless of the actual

levels of in vivo protein prenylation. Analysis of protein

prenylation in vivo is presented in Chapter III.



The testicular cell populations examined in Chapter II

were observed to have developmental changes in PFT and PGGT-

I activities, yet several questions remain unexplored. For

example, does the level of in vivo prenylation of proteins

reflect the changes observed in the in vitro levels of PFT

and PGGT-I activities? Does the mode of prenylation

(farnesyl or geranylgeranyl) reflect the levels of the

farnesyl and geranylgeranyl transferase activities in rats

of different ages and in isolated spermatogenic cells? Does

the ratio of geranylgeranylated to farnesylated proteins

change in testicular development? Do the types of

prenylated proteins change in development and are they

different between spermatogenic cell types? Elucidating the

answers to these questions will be of value in understanding

the regulation of prenylated protein biosynthesis during

spermatogenesis in rat. It is well documented that the

production of isoprenoids is highly regulated in the cell.

Brown and Goldstein have reported several levels of

regulation of mevalonate synthesis in the cell including


_ _

cholesterol feedback regulation of the enzymes HMG CoA

synthase and HMG CoA reductase and the regulation of

expression of LDL receptors. This regulation is deemed

necessary to prevent the overproduction of cholesterol which

is in excess in atherogenic tissues (30). Similar feedback

regulation controls the expressionof other enzymes of the

isoprenoid biosynthetic pathway, including prenyltransferase

(farnesyl diphosphate synthase) (31, 32) and squalene

synthase (30). Cells also regulate the distribution of

mevalonate products by shunting isoprenoids into non-sterol

pathways which has been termed the flux diversion hypothesis

by Faust et al. (106). The flux is regulated by the enzymes

of the non-sterol pathway which have higher affinities for

mevalonate-derived substrates (e.g. FPP) than the enzymes of

the sterol pathway. This feature is particularly important

when mevalonate is limiting (30). These enzymes include PFT,

whose K. for FPP is determined to be 0.04 jM with the

purified enzyme from bovine brain (107) and <0.1 gM with

purified enzyme from rat brain (75). The K. of all-trans-

GGPP synthase for FPP was determined to be 0.6 MM (26),

which is significantly lower than that for cis-

prenyltransferase (Km = 5-24 AM) (29, 108) and squalene

synthase (Km = 1.0 iM) (109). This ensures that sufficient

FPP and GGPP are available for protein modification.

Regulating the availability of isoprenoids for non-

sterol dependent pathways during spermatogenesis was first

described by the work of Potter et al. who observed

independent regulation of dolichol and cholesterol synthesis

in the developing mouse testes (23). The rate of acetate

incorporation into cholesterol increased from the

preleptotene to prepuberal pachytene spermatocytes, then

decreased in late pachytene spermatocytes to remain low

throughout meiosis and in mature sperm. Dolichol synthesis

increased in cells at early stages, remained high in

pachytene spermatocytes and round spermatids, and dropped to

low levels of incorporation in mature sperm. The activity of

HMG CoA reductase mirrors the pattern of cholesterol

synthesis. Therefore, in the developing stages of

spermatogenesis and in isolated spermatogenic cells, even

when HMG CoA reductase levels are low and cholesterol

synthesis is low, there is enough isoprenoid flux for

dolichol synthesis and probably for the prenylation of

proteins in vivo.

The availability of mevinolin, a potent competitive

inhibitor of HMG CoA reductase, has permitted the evaluation

of isoprenoid addition to proteins. Schmidt et al. found

that when radioactive mevalonate was added to the cultures

of mevinolin treated Swiss 3T3 cells, proteins of 13-58 kDa

were labelled with mevalonate metabolic products (48).

Several investigators have noted that concentrations of 1-25

gM mevinolin (also called lovastatin) were sufficient to

inhibit mevalonate synthesis, FPP and GGPP synthesis, and

protein prenylation (66, 110-112). Repko and Maltese

observed that incubating murine erythroleukemia cells with

lovastatin for as little as 1 h prior to the addition of

cycloheximide and [3H)-mevalonic acid rendered the

prenylation of proteins cycloheximide insensitive (110).

Since protein translation was not necessary for protein

prenylation following lovastatin pretreatment, it was

concluded that the blocking of mevalonate synthesis caused a

depletion of isoprenoid groups and allowed an accumulation

of non-isoprenylated substrate proteins that could be

subsequently labelled with radioactive mevalonate precursor.

This protocol also increases the level of protein labelling

above that of non-mevinolin treated cells since depletion of

endogenous mevalonate results in a high specific activity of

the mevalonic acid when labelled mevalonic acid is added.

The work described in this chapter applied the

principle described by Repko and Maltese where mevinolin was

used to block endogenous mevalonate synthesis in primary

cultures of seminiferous epithelium or of isolated

spermatogenic cells before the testicular proteins were

metabolically labelled with [3H]-mevalonic acid. This

chapter shows 1) optimization of incubation times and drug

concentrations for the mevinolin treatment and [3H]-

mevalonic acid labelling in the testes system; 2) the

analysis of the types and proportion of polyprenols released

from (3H]-mevalonic acid labelled proteins by methyl iodide

__ _