lysosomal integral membrane proteins exhibit region and cell type specific distribution in the...

THE ANATOMICAL RECORD 23285-96 (1992)

Lysosomal Integral Membrane Proteins Exhibit Region and Cell Type Specific Distribution in the Epididymis of the Adult Rat

CARLOS A. SUAREZ-QUIAN, NICOLE JELESOFF, AND STEPHEN W. BYERS Georgetown University Medical Center, Department of Anatomy and Cell Biology,

Washington, D.C. 20007

ABSTRACT The epididymis, a post-testicular site required for maturation and storage of spermatozoa, is actively involved in exocytic and endocytic events, two phenomena likely to depend on the integrity of the lysosomal system. To study the lysosomal system of the epididymis, five monoclonal antibodies, previously char- acterized as recognizing five distinct lysosomal integral membrane proteins (LIMPs 1-5), were used as molecular probes of lysosome distribution in cells lining the epithelium. Immunocytochemical localization of LIMPs, using biotin-strepta- vidin immunoperoxidase methodology, was performed on frozen sections of adult rat epididymides and in cell cultures prepared from either the caput or cauda epididymis. In frozen sections, a heterogeneous distribution of the different LIMPs along the length of the epididymis was observed. For example, the distribution of LIMP 1 (35-50 K) was detected in all cells of the caput and quite dramatically in clear cells of the distal caput, corpus, and cauda epididymis, but specifically not in the principal cells of the distal caput, corpus, and cauda. In contrast, LIMP 2 (64-71 K) was present in all cells of the epididymis, except clear cells. LIMPs 4 and 5 (93 K and 93 K) were detected in all epididymal cells, including the clear cells. Finally, whereas the regional and cell type distribution of LIMP 3 (74 K) in the epididymis was identical to that of LIMPs 4 and 5, the nature of the vesicles immunostained was distinct. In cultured cells, the general immunostaining pat- terns observed in vivo were maintained during the duration of the primary cul- tures for all five LIMPs. Our results begin to address the molecular heterogeneity of the lysosomal system along the length of the epidiymis, and may suggest in part a basis for underlying structural and functional characteristics of the epididymis leading to the sequential maturation of sperm.

The functions of the epididymis include storage, transport, and provision of a post-testicular site for the maturation of spermatozoa (Bedford, 1975; Hamilton, 1975; Cooper, 1986; Orgebin-Crist, et al., 1987; Robaire and Hermo, 1988). A means by which the epididymis accomplishes these functions is to generate a special luminal fluid that “conditions” the spermatozoa. The unique composition of this luminal fluid is believed to be the result, in part, of specific secretory and endocytic activity in a region specific manner of the cells lining the ductus epididymis.

Central to the secretory and absorptive activity of epididymal epithelial cells is the Golgi-lysosomal sys- tem. Whilst the extent and structure of these or- ganelles has been investigated by several workers (re- viewed in Robaire and Hermo, 1988) the molecular makeup of epididymal lysosomes and Golgi is un- known. In the present study we use five monoclonal antibodies that recognize five distinct lysosomal inte- gral membrane proteins (LIMPs) to begin the molecu- lar characterization of the epididymal lysosomal sys- tem. Specifically, we determined the distribution of the five LIMPs by biotin-streptavidin immunoperoxidase methodology in different regions of the epididymis and in cultured epididymal epithelial cells. Our results

0 1992 WILEY-LISS, INC.

demonstrate that the previously described regional heterogeneity in exocytic andlor endocytic events along the length of this duct is accompanied by variations in the distribution of LIMPs.

MATERIALS AND METHODS Monoclonal Antibodies

Monoclonal antibodies were generated against par- tially purified hepatocyte, lysosomal membrane pro- teins (Leighton et al., 1968) devoid of peripheral pro- teins by treatment with 0.1 M Na2C0, (Fujiki et al., 1982). The production of hybridomas secreting antibod- ies recognizing LIMPs and the biochemical properties of the LIMPs themselves have been described in detail previously (Barriocanal et al., 1986; Bonifacino et al., 1986; Suarez-Quian, 1987, 1988). LIMPs 1-5 are all glycoproteins, are recognized by monoclonal antibodies 14E12, 5G10, 29G10, 21D7, and 38C7, respectively, and their respective kilodaltons on sodium dodecyl sul- fate (SDS)-polyacrylamide gels are approximately 35- 50,64-71,74,93, and 93. The specific kilodaltons were

Received March 14, 1991; accepted June 28, 1991

86 C.A. SUAREZ-QUIAN ET AL.

Fig. 1. Control studies. A A positive method control for the biotin- streptavidin immunoperoxidase technique using antibody 37B3 that specifically immunostains the nuclear lamins of somatic cells is shown. Frozen sections of epididymis were prepared and immuno- stained as described in the Materials and Methods Section. Note that the cytoplasm of the epididymal principal cells is not immunostained with the antibody, nor the nuclei of spermatozoa present within the

lumen of the epididymis. Cell nuclei in the interstitium are also spe- cifically immunostained with the 37B3 antibody, but in addition, non- specific peroxidase reaction product is observed here. B: A profile of a negative control section, in which primary antibody was omitted, is illustrated. Nuclei (N) were counterstained with hematoxylin. Note that non-specific peroxidase reaction product is present in the inter- stitium also. A,B: x 878.

EPIDIDYMAL LYSOSOMAL SYSTEM 87

Fig. 2. Distribution of LIMPs in the proximal caput. Immunocy- tochemical localization of LIMP 1 (A,A’), LIMP 2 (B,B’), and LIMPs 4-5 (D,D’) in the proximal caput is demonstrated. Arrows point to vesicles immunostained with the antibodies. Nuclei (N) were coun- terstained with hematoxylin. Note that specific luminal immuno- staining of LIMP 1 (A,A‘) is detected in the proximal caput and that the number of cytoplasmic vesicles immunostained is fewer for LIMP 1 than for LIMPs 2, 4, and 5. LIMPs 4 and 5 immunostaining was

detected in the infranuclear, supranuclear, and apical cytoplasm of principal cells, whereas only supranuclear and apical cytoplasmic im- munostaining was detected for LIMPs 1, 2, and 3. C: LIMP 3 immu- nostaining at the junction of efferent ductules (ED) and its absence in the initial segment (IS) of the proximal caput is demonstrated. C’: A higher magnification image of the efferent ductules immunostaining pattern is shown. A-D, X 268; A’-D’, X 878.

88 C.A. SUAREZ-QUIAN ET AL.

TABLE 1. Summary of biochemical properties of the five LIMPs derived from rat liver and normal rat kidney cells

obtained by from either

Molecular weight Lysosome Endosome Cell surface

Antibody Protein (kD) staining staining staining 14E12 LIMP 1 35-50 + + +

LIMP 2 64-71 + + + 5G10 29G10 LIMP 3 74 + ~

21D7 LIMP 4 93 + + + 38C7 LIMP 5 93 + + (Data from Barriocanal et al., 1986; Suarez-Quian, 1987, 1988.)

-

-

immunoprecipitating solubilized proteins rat livers or normal rat kidnev cells that

had been radiolabeled with lZ5I (Suarez-Qkan, 1987, 1988). The only antibody that works in Western blot- ting is the 29G10; thus, this is not a viable method to study the molecular size of the five LIMPs. A summary of the distinct properties of the different LIMPs is pre- sented in Table 1.

Approximately 1 1 from each hybridoma culture was precipitated at a 1:l ratio with saturated ammonium sulphate at 4"C, brought up to a 20 x concentration in phosphate-buffered saline (PBS), and dialyzed exten- sively against PBS. This stock solution of monoclonal antibody was diluted 1500 and 1:1,000 and used in the immunocytochemical studies (see below).

lmmunocytochemistry Four adult Sprague-Dawley rats were killed by CO,

narcosis, and the epididymides were removed quickly and frozen in liquid nitrogen. The methodology em- ployed for freezing (liquid NJ, sectioning (5 pm frozen sections), and fixing (formalin or ice cold methanol) the epididymis was developed empirically first in the liver, the tissue from which lysosomes were purified and used to raise the monoclonal antibodies (Barriocanal et al., 1986). In addition, it has been used with considerable success previously to immunolocalize androgen bind- ing protein in the epididymis (Pelliniemi et al., 1981). Next, the epididymides were placed in a 2800 Frigocut (Reichert-Jung; Baltimore, MD) Cryostat a t - 15°C and left to reach equilibrium for 2 hours. Sections were cut a t a thickness of 5 microns and fixed for 10 minutes with either 3.7% formalin in PBS, or 5 minutes in ice cold methanol. Histostain-SP kits (Zymed Laborato- ries; San Francisco, CA) for mouse primary antibodies were used exactly as described by the manufacturer's instructions to determine the precise epididymal dis- tribution of the lysosomal integral membrane proteins. These kits employ the biotin-streptavidin peroxidase system to illustrate the positive, cellular immunoreac- tion product. In this method a positive signal is iden- tified by a reddish-brown reaction product. The proce- dure is as follows: the endogenous peroxidase activity is blocked by using a periodic acid solution (Zymed) by adding one drop to each cryostat section and incubating for exactly 45 seconds (Kelly et al., 1987). The immu- nocytochemical procedure involved five steps: 1) sec- tions were incubated in normal rabbit serum blocking solution for 10 minutes; 2) the mouse monoclonal an- tibodies were added and sections were incubated for 1 hour a t 37°C in a moist chamber followed by extensive

washes in PBS; 3) biotinylated second antibody (rabbit anti mouse) was added to each section and incubated for 30 minutes a t 37°C in a moist chamber followed by extensive washes with PBS; 4) the enzyme conjugate, streptavidin-peroxidase, was then added and sections were incubated at room temperature for 5 minutes fol- lowed by extensive washing with PBS; 5) next, the sub- strate-chromogen mixture, H,O,-aminoethyl carba- zole, was added to each section and incubated for 15 minutes at room temperature followed by extensive washing with H,O.

Control immunocytochemical studies employed in- cluded both negative (steps 1-4 below) and positive (step 5) method controls as follows: 1) omission of mouse monoclonal antibody; 2) conditioned media from a hybridoma that do not give rise to a positive reaction by immunocytochemistry; 3) normal mouse sera were used; 4) hybridoma conditioned media depleted of an- tibodies by running through a Sepharose-protein G col- umn (Pharmacia) were used; 5 ) as a positive method control for immunocytochemistry, another monoclonal antibody was used, 37B3 (Suarez-Quian, 1987, 1988), that recognizes a nuclear lamina protein.

After the immunocytochemical reaction, frozen sec- tions were counterstained with hematoxylin. To maxi- mize visualization of the red reaction product, frozen sections were examined with a Zeiss, Planapo 63x phase 3, 1.4 N.A. objective using an 80A blue filter. Images were recorded on Kodacolor 100 ASA film and processed by a professional photographic laboratory (CPI, Bethesda, MD).

Epididymal Cell Isolation and Culture Epididymal epithelial cells were isolated from either

the caput-corpus or cauda epididymides of 40-50 day old rats. The epididymides were divided at the isthmus of the corpus epididymides. After a single 60-90 minute collagenase digestion (Byers et al., 1985,1986) free tubule fragments were incubated with 0.2% pro- tease in calcium-magnesium free Hanks balanced salt solution to release sheets of epithelial cells (Byers et al., 1988). Cells were grown on glass cover slips in se- rum free defined medium as described earlier (Byers et al., 1985, 1986).

Electron Microscopy Cells were prepared for electron microscopic exami-

nation as described previously (Suarez-Quian, 1988; Byers et al., 1985,1986). Cells were fixed with 5% glu- taraldehyde in 0.2 M s-Collidine, post-fixed in 1% os- mium tetroxide, dehydrated through a graded series of

EPIDIDYMAL LYSOSOMAL SYSTEM 89

alcohols, and embedded in Epon. Sixty to 90 nm sec- tions were cut with an LKB microtome and photomi- crographs prepared with a JEOL 100s electron micro- scope.

RESULTS LIMPS lmmunolocalization in Different Regions

of the Epididymis In positive method controls specific immunostaining

was observed within the nucleus of cells lining the duc- tus epididymis, not in the cytoplasni, and non-specific staining was detected in the interstitium (Fig. 1A). In negative control sections, non-specific reaction product was observed only in the interstitium (Fig. 1B). Be- cause treatment of tissue with periodic acid completely eliminated the endogenous peroxidase activity only of the epithelial cells lining the ductus epididymis, we did not examine the lysosomal system of the cells found in the interstitium. These studies ascertained the validity of the biotin-streptavidin immunoperoxidase technique as a means of studying the lysosomal system of cells forming the epithelium of the epididymis.

The distribution of the five LIMPs within the different regions of the epididymis examined in this study and identified by the respective five monoclonal antibodies is demonstrated in Figures 2-4 and is sum- marized in Table 2. In general, qualitative differences were detected in immunostaining intensity and pat- terns when the different monoclonal antibodies were employed as probes of the lysosomes, except between LIMPs 4 and 5. In addition, the vesicular nature of the immunostaining, i.e., “doughnut-like,” could be dis- cerned readily in favorable profiles; this pattern is consistent with what is known about the five LIMPs, mainly that they are resident proteins of the mem- brane container of lysosomes. Absence of “doughnut- like” patterns in positive cells was interpreted as resulting from superimposition of several immuno- stained vesicles within the 5 pm frozen sections, rather than from poor fixation.

LIMP Distribution in the Proximal Regions of the Epididymis

In the initial segment of the epididymis numerous vesicles were immunostained with the anti LIMP 1, 2, 4, and 5 antibodies and the pattern of immunostaining was quite similar to that observed in the proximal caput (data not shown). In contrast, no LIMP 3 immu- nostaining was detected in the initial segment, whereas very strong staining was observed in the ef- ferent ductules (Fig. 2 0 .

In the proximal caput the lysosomes containing LIMP 1 were distributed throughout the height of the epithelium, whereas vesicles immunostaining for LIMP 3 were localized in a more perinuclear area. In addition, prominent LIMP 1 immunostaining, but not LIMP 3 staining, was detected in the lumen of the caput epididymis (Fig. 2A,A’). Many more lysosomes containing LIMPS 2,4 , and 5 were observed than those expressing LIMPs 1 and 3. Subtle differences were also detected between lysosomes expressing LIMP 2 vs LIMPs 4 and 5. For example, more vesicles in the peri- nuclear Golgi region were immunostained for LIMP 4 and the number of lysosomes immunostained with anti

LIMPs 4 and 5 were greater than those identified with anti LIMP 2.

LIMP Distribution in the Distal Caput Epididymides Immunocytochemical localization of LIMPs 1-5 in

the distal caput epididymides is presented in Figure 3. Specific LIMP 1 immunostaining in this region was observed in clear cells and not in principal cells. The identity of clear cells was ascertained by the appear- ance of large vacuoles within the cytoplasm. It is in this region of the epididymis that clear cells are first de- tected (Robaire and .Hermo, 1988). Luminal LIMP 1 immunostaining was also detected, although the inten- sity was much reduced when compared to the intensity of luminal LIMP 1 immunostaining in the proximal caput. In contrast, LIMP 2 immunostaining was de- tected only in principal cells of the distal caput and not in the clear cells (Fig. 3B,B’). In addition, LIMP 2 im- munostaining appeared to be predominantly located in the supranuclear and apical cytoplasm of the principal cells.

Neither LIMP 3 (Fig. 3C,C’) nor LIMPs 4-5 (Fig. 3D,D’) immunostaining in the distal caput appears to be limited to specific cell types of this region. All sec- tions of the distal caput examined for LIMPs 3-5 re- vealed that these proteins were present in all cell types. However, there were marked differences in the immunostaining characteristics between LIMP 3 and LIMPs 4-5. Significantly, LIMP 3 immunostaining was limited to a supranuclear region of cells, whereas LIMPs 4 and 5 immunostaining revealed clearly a su- pranuclear, infranuclear, and an apical cytoplasmic disposition of the proteins. In addition, the number of vesicles immunostained for LIMP 3 was less than the number of vesicles immunostained for the other four LIMPs.

LfMP Distribution in the Cauda Epididymides The distribution of LIMPs 1-5 in the cauda epididy-

mis is demonstrated in Figure 4. No differences were discerned between the corpus and cauda epididymides in the immunostaining patterns of LIMPs 1-5 and thus only results of the latter region are presented here. As in the distal caput, LIMP 1 immunostaining appeared to be limited to clear cells of the cauda epididymis (Fig. 4A,A’). In this region, the number of clear cells in- creases to 12% (Robaire and Hermo, 19881, and, al- though quantitative values of LIMP l immunostaining in clear cells were not obtained, qualitatively many more clear cells were specifically immunostained for LIMP 1 in this region. In addition, the luminal LIMP 1 immunostaining appeared to be limited to the region directly above clear cells (Fig. 4A).

The LIMP 2 immunostaining pattern observed in the distal caput was maintained in the cauda epididymis (Fig. 4B,B’). Principal cells were immunostained spe- cifically for LIMP 2, whereas clear cells were not. As in the distal caput, LIMP 2 immunostaining within the principal cells appeared to be limited to the supranu- clear cytoplasm with occasional infranuclear staining.

The immunostaining patterns of LIMP 3 (Fig. 4C,C‘) and LIMPs 4-5 (Fig. 4D,D‘), respectively, observed in the cauda epididymis were similar to the patterns ob- served for the three LIMPs in the distal caput. As in the distal caput, LIMP 3 appeared limited to the SU-

90 C.A. SUAREZ-QUIAN ET AL.

Fig. 3. Distribution of LIMPs in the distal caput. Immunolocaliza- tion of LIMP 1 (A,A’), LIMP 2 (B,B’), LIMP 3 (C,C’), and LIMPs 4-5 (D,D’) in the distal caput is demonstrated. Nuclei (N) were counter- stained with hematoxylin. Specific LIMP 1 immunostaining is present in clear cells (curved arrows) in the lumen, and to a lesser extent, if at all, in principal cells. LIMP 2 immunostaining is present in principal cells but does not appear in clear cells (curved arrows). Clear vacuoles (arrowheads) in A’ and B’, characteristic of clear

cells, are labeled. LIMP 3 immunostaining is confined to the supra- nuclear cytoplasm and the number of vesicles immunostained for LIMP 3 is significantly less than for the other LIMPs. LIMPs 4 and 5 immunostaining was present in all cells of the distal caput, and within the cells, the immunostaining was detected in the infranu- clear, supranuclear, and apical cytoplasm. Samples of vesicle profiles immunostained for LIMPs (short arrows) are labeled in C’ and D‘. A-D, x 268; AI-D’, x am.

EPIDIDYMAL LYSOSOMAL SYSTEM 91

Fig. 4. Distribution of LIMPs in the cauda epididymis. Immunolocal- ization of LIMP 1 (A,A’), LIMP 2 (B,B’), LIMP 3 (C,C’), and LIMPs 4-5 (D,D’) in the cauda is demonstrated. Clear cells (curved arrows) are immunostained specifically for LIMP 1, but are not immuno- stained for LIMP 2. In contrast, principal cells are immunostained for LIMP 2 but not LIMP 1. Nuclei (n) were counterstained with hema- toxylin. Slight luminal LIMP 1 immunostaining can be discerned in this region of the epididymis. White arrows (A’) and arrowheads (B’)

indicate vacuoles present within clear cells. LIMP 2 immunostaining appears limited to the supranuclear and apical cytoplasm of principal cells. LIMP 3 immunostaining is limited to the perinuclear cytoplasm and the number of vesicles immunostained for LIMP 3 is significantly less than for the other LIMPs. LIMPs 4 and 5 immunostaining was detected in all cells of this region and was present in the infranuclear, supranuclear, and apical cytoplasm. Arrows in C’ and D’ indicate samples of vesicles immunostained for LIMPs.

92 C.A. SUAREZ-QUIAN ET AL.

TABLE 2. Cell and regional distribution of LIMPs in epididyrnis'

Distal caput Cauda Initial Proximal Principal Clear Principal Clear

Protein segment camt cells cells cells cells + - + - + - + -

LIMP1 + LIMP2 + + + LIMP3 - + + + + + LIMP4 + + + + + + 1,TMP 5 + + + + + + 'For more details see the Results section

pranuclear cytoplasm, whereas LIMPs 4-5 were de- tected throughout the infranuclear, supranuclear, and apical cytoplasm of cells. In addition, LIMP 3 appeared to be present in much fewer vesicles than the other LIMPs.

LIMPS lmmunolocalization in Cultured Epididymal Cells The principal advantage of performing LIMP immu-

nocytochemistry in epididymal cells in vitro is that the structure of the vesicles can be better discerned at the level of resolution provided by the light microscope than in frozen sections. It also provides information on the maintenance of differentiated characteristics of cultured cells. Results revealed that the number of ves- icles immunostained for LIMP 5 (Fig. 5F) was greater than the number of vesicles immunostained for LIMP 4 (Fig. 5E) and LIMP 2 (Fig. 5C). In addition, a diffuse LIMP 4 immunostaining pattern in the perinuclear Golgi region was often discerned in these cells.

The number of vesicles immunostained for LIMP 3 (Fig. 5D) was less than the number of vesicles immu- nostained for the other LIMPs. Furthermore, the dis- tribution of the vesicles immunostained for LIMP 3 appeared limited to the perinuclear cytoplasm and were irregular in shape, exhibiting tubular append- ages.

In mixed populations of epididymal cells in vitro clear cells can be readily identified by their abun- dances of vesicles (Fig. 6A). As reported above, the dis- tribution of LIMP 1 in the distal caput, corpus, and cauda epididymis was limited to clear cells. In epidid- ymal cells in vitro, prepared from corpus and cauda epididymides, LIMP 1 immunostaining was also lim- ited to clear cells (Fig. 6B) rendering these cells readily identifiable in the immunostained preparations.

DISCUSSION In this study we describe the distribution of five dis-

tinct lysosomal integral membrane proteins in differ- ent regions of the epididymis. Our principal finding is that there is a regional and cell specific expression of five different LIMPs in the epididiymis. These results are summarized in Table 2. Our observations suggest that, even at the organelle level, distinct regions of the epididymis manifest different protein complementa- tion. In a subsequent manuscript we will describe sim- ilar regional and cell type differences in the molecular composition of the epididymal Golgi complex. We wish to emphasize, however, that our methodology was not sensitive enough to discern potential cell specific ex- pression of LIMPs in other cell types, e.g., basal, nar-

row, or halo cells, of the epididymis. In cultured epi- didymal cells, the respective LIMP immunostaining patterns mimicked the patterns observed in the epidid- ymis in vivo, and significantly, the immunoreaction was vesicular and was present as a rim outlining the membrane container of the vesicles.

The demonstration that the five LIMPs are ex- pressed in a region and cell specific fashion along the epididymis may underscore vital regional differences in endocytic activity in this tissue. Regional differences have also been described for secreted proteins (Kohane et al., 1980; Flickinger, 1981; Brooks and Tiver, 1984; Klinefelter and Hamilton, 1985; Olson and Hinton, 1985; Cornwall et al., 1990; Vreeburg et al., 1990; Hermo et al., 1991) and plasma membrane antigens (Byers et al., 1988; Byers and Graham, 1990). Simi- larly, there is evidence in the literature that one other membrane protein of the endocytic pathway is ex- pressed in a region specific fashion in an absorptive epithelium. A 55-61 kD protein was found exclusively in the apical endocytic complex in the ileal absorptive cells of suckling rats (Wilson et al., 1987) but not in other regions of the gut. Interestingly, the ultrastruc- ture of equivalent structures resembling lysosomes varies in caput, corpus, and cauda (Robaire and Hermo, 1988). From the proximal to distal epididymis, the ly- sosomes become more electron dense. Presumably, this change in the ultrastructure of lysosomes is a function of differences in uptake of different substances along the length of the epididymis and may require distinct protein components in the membrane containers of ly- sosomes.

A possible explanation of the immunolocalization of LIMP 1 in clear cells is that it represents endocytosed LIMP 1. Consistent with this is the fact that clear cells may be quite active in phagocytosis as evident by the uptake of the released cytoplasmic droplet of sperma- tozoa (Hermo et al., 1988). However, in epididymal cells in vitro LIMP 1 was also immunolocalized in ves- icles within clear cells, making this interpretation un- likely. Nevertheless, the luminal staining of an inte- gral membrane protein found predominantly in lysosomes is puzzling and needs to be addressed: lyso- soma1 hydrolases are commonly released into the ex- tracellular environment. Similarly, it should be men- tioned that up to 12% of total cellular LIMP 1 is found within the plasma membrane of normal rat kidney cells at steady-state (Suarez-Quian, 1988). Hence, it is not surprising that a portion of LIMP 1 is accessible for release into the lumen of the epididymis. Given that LIMP 1 may be detected a t the cell surface and possibly secreted, what are the potential mechanisms of this event? Evidence that LIMP 1 is an integral membrane protein comes from detergent partitioning studies us- ing triton X-100 (Barriocanal et al., 1986; Suarez- Quian, 1988). Thus, one possibility that has not been examined rigorously is that LIMP 1 is attached to the membrane via a phosphatidyl linkage and as such may be released into the lumen by a phospholipase. Alter- natively, LIMP 1 may truly be an integral membrane protein, but with only one membrane spanning do- main, and be released readily into the lumen by prote- olysis. In this context, a similar observation has been made for the enzyme gamma-glutamyl transpeptidase which is an integral membrane protein of the caput

EPIDIDYMAL LYSOSOMAL SYSTEM 93

epididymis and is proteolytically released into the lu- men, where it is found in membrane bound and a non- membrane bound form (Hinton et al., 1990). In addi- tion, certain monoclonal antibodies raised against epididymal plasma membranes also stain determi- nants present in the lumen (Byers et al., 1988). What- ever the true nature of the LIMP 1 interaction with the membrane domain is, the apparent “secretion” of these two plasma membrane proteins of the caput ep- ididymis may represent a general phenomenon of the epididymis in this region and should be examined more closely.

The epididymis is traditionally subdivided into the initial segment, caput, corpus, and cauda based on re- gional changes in the histology (Robaire and Hermo, 1988). Our results with LIMP 1 suggest also that a functional difference exists between the principal cells of the proximal vs. distal caput. That is, whereas in the proximal caput the principal cells immunostain for LIMP 1, in the distal caput, i.e., the region where clear cells first appear, no LIMP 1 was detected. Although the significance of the LIMP 1 distribution is not clear, it does suggest that even between morphologically sim- ilar principal cells regional differences in function may exist. A similar conclusion has been reached recently by Hermo et al. (1991) with respect to the uptake of SGP-2 in the epididymis.

As membrane markers of cytoplasmic vesicles that take part in endocytosis have been identified, the ly- sosome as a structure has been redefined as exhibiting lgp 120 (the homologue of LIMP 4) immunostaining but lacking the mannose-6-phosphate receptor (the receptor which mediates the transport of lysosomal en- zymes from their site of synthesis to lysosomes) (Korn- feld and Mellman, 1989). Based on previous character- ization studies, LIMPs 1-5 are found predominantly in lysosomes. However, LIMPs 1 ,2 , and 4 are also present in different vesicles of the endocytic pathway (Suarez- Quian, 1987, 1988). That is, LIMPs 1, 2, and 4 can be promiscuous residents of the membrane container of vesicles involved in uptake. Interestingly, in the epi- didymis only LIMP 3 appeared to be a resident exclu- sively of vesicles which encompass the classical, as well as the more recent, definition of lysosomes. These ves- icles were few in number and observed in a supranu- clear position in vivo and in a perinuclear area in vitro. In contrast, our results suggest that LIMPs, 1 ,2 ,4 , and 5 probably inhabit the endosomal vesicles as well as the lysosomes. In addition, a distinction must be made between the different vesicles immunostained for LIMPs 2,4, and 5 and those which contain only LIMPs 4 and 5. LIMP 2 staining was not detected in the in- franuclear cytoplasm of principal cells, whereas LIMPS 4 and 5 staining was readily observed in this region. This distinction in intracytoplasmic vesicles based on the immunostaining patterns may suggest a mecha- nism by which the principal cells segregate the apical cytoplasmic compartment, which presumably responds to endocytosed material from the lumen, from the basal cytoplasmic compartment, presumably responding to extracellular signals from blood. Whereas in a non- polarized cell this apportionment of vesicles may not be important, in a polarized cell such domain specific seg- regation of vesicle contents may be vital to its normal activity.

ACKNOWLEDGMENTS This work was supported in part by NIH grant

HD23484 to C.A.S.-Q. and NIH grants HD25028 and HD23744 to S.W.B.

LITERATURE CITED Barriocanal, J.G., J.S. Bonifacino, L. Yuan, and I.V. Sandoval 1986

Biosynthesis, glycosylation, movement through the Golgi system and transport to lysosomes by a N-linked carbohydrate indepen- dent mechanism of three lysosomal integral membrane proteins. J . Biol. Chem., 261t16755-16763.

Bedford, J.M. 1975 Maturation, transport and fate of spermatozoa in the epididymis. In: Handbook of Physiology, Section 7: Endo- crinol. R.O. Greep and E.B. Astwood, eds. American Physiological Society, Washington, D.C., Vol. 5, pp. 303-317.

Bonifacino, J.S., P. Perez, R.D. Klausner, and I.V. Sandoval 1986 Study of the transit of a n integral membrane protein from secre- tory granules through the plasma membrane of secreting rat ba- sophilic leukemia cells using a specific monoclonal antibody. J. Cell Biol., 102:516-522.

Brooks, D.E., and K. Tiver 1984 Analysis of surface proteins on rat spermatozoa during epididymal transit and identification of an- tigens common to spermatozoa, rete testis fluid and cauda epi- didymal plasma. J . Reprod. Fertil., 71t983-988.

Byers, S.W., D. Djakiew, and M. Dym 1985 Structural characteristics of epididymal epithelial cells in vitro. J. Reprod. Fertil., 75t401- 411.

Byers, S.W., and R. Graham 1990 Distribution of sodium-potassium ATPase in the rat testis and epididymis. Am. J . Anat., 188:31- 43.

Byers, S.W., M.A. Hadley, D. Djakiew, and M. Dym 1986 Growth and characterization of polarized monolayers of epididymal epithelial cells and Sertoli cells in dual environment culture chambers. J. Androl., 7.59-68.

Byers, S.W., L. Richardson, and C. Parker 1988 Isolation and charac- terization of epididvmal epithelial cell plasma membranes. Biol. Reprod., 38:201-209. ~

Cooper, T.G. 1986 The Epididymis, Sperm Maturation and Fertilisa- tion. Springer-Verlag, New York, pp. 1-281.

Cornwall, G.A., J.T. Vreeburg, M.K. Holland, and M.X. Orgebin- Crist 1990 Interactions of labeled epididymal secretory proteins with spermatozoa with injection of ”S-methionine in the mouse. Biol. Reprod., 43:121-129.

Flickinger, C.J. 1981 Regional differences in synthesis, intracellular transport, and secretion of protein in the mouse epididymis. Biol. Reprod., 25.871-873.

Fujiki, Y., A.L. Hubbard, S. Fowler, and P.B. Lazarow 1982 Isolation of intracellular membranes by means of sodium carbonate treat- ment; application to endoplasmic reticulum. J. Cell Biol., 93:97- 102.

Hamilton, D.W. 1975 Structure and function of the epithelium lining the ductuli efferentes, ductus epididymis and ductus deferens in the rat. In: Handbook of Physiology, Section 7: Endocrinol. R.O. Greep and E.B. Astwood, eds. American Physiological Society, Washington, D.C., Vol. 5, pp. 259-301.

Hermo, L., J. Dworkin, and R. Oko 1988 Role of epithelial clear cells of the rat epididymis in the disposal of the contents of cytoplasmic droplets detached from spermatozoa. Am. J . Anat., 183.107-124.

Hermo, L., J . Wright, R. Oko, and C.R. Morales 1991 Role of epithelial cells of the male excurrent duct system of the rat in the endocy- tosis of secretion of sulfated glycoprotein-2 (clusterin). Biol. Re- prod., 44:1113-1131.

Hinton, B.T., M.A. Palladino, D.R. Mattmueller, D. Bard, and K. Good 1991 Expression and activity of gamma-glutamyl transpeptidase in the rat epididymis. Mol. Reprod. Dev., 1.40-46.

Kelly, J . , C.A. Whelan, D.G. Weir, and C. Feighery 1987 Removal of endogenous peroxidase activity from cryostat sections for immu- noperoxidase visualization of monoclonal antibodies. J. Immunol. Methods, 96127-132.

Klinefelter, G.R., and D.W. Hamilton 1985 Synthesis and secretion of proteins by perifused caput epididymal tubules, and association of secreted proteins with spermatozoa. Biol. Reprod., 33.1017- 1027.

Kohane, A.C., M.S. Cameo, L. Pineiro, and J.A. Blaquier 1980 Dis- tribution and site of production of specific proteins in the rat epididvmis. Biol. Rearod.. 23,181-187.

Kornield, S., and I. Mellkan 1989 The biogenesis of lysosomes. Ann.

Leighton, F., B. Poole, H. Beaufay, P. Baudhuin, J.W. Coffey, S. Rev. Cell Biol., 5:483-525.

94 C.A. SUAREZ-QUIAN ET AL.

Fig. 5. Immunocytochemical localization of LIMPs in epididymal cells in vitro. Immunolocalization of LIMP 2 (C), LIMP 3 (D), LIMP 4 (El, and LIMP 5 (F) in primary cultures of epididymal cells is shown. A positive method control, using a monoclonal antibody to a nuclear lamin, is shown in A and a negative control, in which the primary antibody was omitted, is shown in B. Nuclei (N) were counterstained with hematoxylin. Note that the vesicles immunostained for LIMP 2 (C) and LIMP 5 (F) may be found close to the plasma membrane

(arrowheads) and that at the light microscopic level there does not appear to be structural differences between the vesicles immuno- stained for either of the two LIMPs. The vesicles immunostained for LIMP 3 (D) are fewer in number and exhibit tubular outpocketings (curved arrow). In E, the perinuclear area, often occupied by the Golgi apparatus, also immunostains for LIMP 4, as well as vesicles in the peripheral cytoplasm. A-F, x 878.

EPIDIDYMAL LYSOSOMAL SYSTEM 95

Fig. 6. Immunolocalization of LIMP 1 in clear cells in vitro. A: A profile of epididymal cells in vitro at the ultrastructural level is il- lustrated. Clear cells can be discerned by the large number of vesicles present in the cytoplasm. B: LIMP 1 immunostaining of a clear cell in

vitro ( C ) from a preparation of cells obtained from the corpus and cauda epididymis is demonstrated. Nuclei of principal cells (PC) are labeled. B, x 878.

96 C.A. SUAREZ-QUIAN ET AL.

Fowler, and C. de Duve 1968 The large-scale separation of per- oxisomes, mitochondria, and lysosomes from the livers of rats injected with Triton WR-1339. J. Cell Biol. 37t482-513.

Olson, G.E., and B.T. Hinton 1985 Regional differences in luminal fluid polypeptides of the rat testis and epididymis revealed by two-dimensional gel electrophoresis. J . Androl., 6~20-34.

Orgebin-Crist, M.-C., L.H. Hoffman, G.E. Olson, and M.D. Skudlarek 1987 Secretion of proteins and glycoproteins by perfused rabbit corpus epididymal tubules: effect of castration. Am. J . Anat., 180: 49-68.

Pelliniemi, L.J., M. Dym, G.L. Gunsalus, N.A. Musto, C.W. Bardin, and D.W. Fawcett 1981 Immunocytochemical localization of an- drogen-binding protein in the male reproductive tract. Endocri- nology, 108t925-931.

Robaire, B., and L. Hermo 1988 Efferent ducts, epididymis, and vas

deferens: structure, functions, and their regulation. In: The Phys- iology of Reproduction. E. Knobil and J. Neill, eds. Raven Press, New York, pp. 999-1080,

Suarez-Quian, C.A. 1987 The distribution of four lysosomal integral membrane proteins (LIMPS) in rat basophilic leukemia cells. Tis- sue Cell, 19:495-504.

Suarez-Quian, C.A. 1988 Differential cell surface expression of four lysosomal integral membrane proteins (LIMPS). Tissue Cell, 20: 35-46.

Vreeburg, J.T.M., M.K. Holland, G.A. Cornwall, and M.-C. Orgebin- Crist 1990 Secretion and transport of mouse epididymal proteins after injection of 35S-methionine. Biol. Reprod., 43:113-120.

Wilson, J.M., J.A. Whitney, and M.R. Neutra 1987 Identification of an endosomal antigen specific to absorptive cells of suckling rat il- eum. J. Cell Biol., 105:691-703.