Investigation of the Role of Prolactin in the Development and Function of the Lacrimal and Harderian Glands Using Genetically Modified Mice Kathleen A. McClellan,1,2 Fiona G. Robertson,3 Jon Kindblom,4 Håkan Wennbo,4 Jan Törnell,4 Brigitte Bouchard,5 Paul A. Kelly,5 and Christopher J. Ormandy 3 PURPOSE. To determine whether prolactin receptor is essential for normal development and function of the lacrimal gland and whether hyperprolactinemia can alter lacrimal development. METHODS. Lacrimal gland morphology and function were examined in two genetic mouse models of prolactin action: a prolactin receptor knockout model that is devoid of prolactin action and a transgenic model of hyperprolactinemia. RESULTS. Image analysis of lacrimal and Harderian gland sections was used to quantify glandular morphology. In females, lacrimal acinar area decreased by 30% and acinar cell density increased by 25% over control subjects in prolactin transgenic animals, but prolactin receptor knockout mice showed no changes. In males, transgenic animals showed no changes, but prolactin receptor knockout mice showed a 5% reduction in acinar area and an 11% increase in acinar cell density, which was lost after castration. The morphology of the Harderian glands underwent parallel changes but to a lesser degree. A complete loss of porphyrin accretions was seen in the Harderian glands of male and female knockout animals. No differences in tear protein levels were seen in knockout animals by two-dimensional gels. Enzyme-linked immunosorbent assay (ELISA) and Western blot analysis showed that the level of secretory component and IgA in knockout mouse tears remained unchanged. There was no change in the predisposition of the 129 mouse strain to conjunctivitis in the knockout animals. CONCLUSIONS. Prolactin plays a small role in establishing the sexual dimorphism of male lacrimal glands. In females, hyperprolactinemia causes a hyperfemale morphology, suggesting a role in dry eye syndromes. Prolactin is required for porphyrin secretion by the Harderian gland but plays no essential role in the secretory immune function of the lacrimal gland. (Invest Ophthalmol Vis Sci. 2001;42:23–30) From the 1Department of Ophthalmology, University of Sydney, Sydney Eye Hospital, Australia; the 3Cancer Research Program, Garvan Institute of Medical Research, Sydney, Australia; the 4Department of Physiology, Research Centre for Endocrinology and Metabolism, Göteborg University, Sweden; and 5Institut National de la Santé et de la Recherche Médicale (INSERM), Faculté de Médecine Necker-Enfants Malades, Paris, France. 2 Present affiliation: Department of Ophthalmology, University of Texas Southwestern Medical Center at Dallas, Texas. Supported by the Ophthalmic Research Institute of Australia (KAM, CJO), the National Health and Medical Research Council of Australia and the New South Wales Cancer Council (CJO), INSERM France (BB, PAK), and the Swedish Cancer Foundation (HW, JT). Submitted for publication July 5, 2000; revised September 11, 2000; accepted October 6, 2000. Commercial relationships policy: N. Corresponding author: Christopher J. Ormandy, Cancer Research Program, Garvan Institute of Medical Research, 384 Victoria Street, Darlinghurst, Sydney 2010, Australia. c.ormandy@garvan.unsw.edu.au Investigative Ophthalmology & Visual Science, January 2001, Vol. 42, No. 1 Copyright © Association for Research in Vision and Ophthalmology I n women, dry eye syndromes occur mostly during alteration in the endocrine environment caused by pregnancy, lactation, oral contraceptive use, or menopause, with consequences ranging from discomfort and contact lens intolerance to persistent pain and corneal damage leading to blindness.1,2 The cause of the disease lies in the disruption of the stability of the tear film, causing poor lubrication between eyelid and globe, resulting in mechanical and inflammatory damage to the corneal epithelium. Disruption of the tear film can be caused by changes in tear composition, supply, drainage, or evaporation and results from deficiencies in the external adnexa including low or excessive tear or tear protein production by the lacrimal gland, poor production of oils by the meibomian gland, low mucus production by the conjunctival goblet cells, and/or abnormal drainage through the tear duct. Adverse environmental conditions such as low humidity, high temperature or high dust levels can exacerbate these deficiencies.3 Current routine treatment relies on artificial tear supplementation or surgical intervention to reduce tear drainage through the canaliculi. Tear deficiency due to declining lacrimal gland function is the major cause of tear film instability and dry eye.4 It carries the additional problem of reduced secretory immunity, because the lacrimal gland is the major source of IgA in tears.5 Two major causes have been identified: primary tear deficiency, the most prevalent cause of dry eye, which results from lacrimal gland destruction by a round cell infiltrate,4 and Sjögren’s syndrome, an autoimmune disease resulting in lymphocytic invasion and destruction of the epithelium of the lacrimal gland.6 The cause of primary tear deficiency remains unknown, but age-related endocrine changes have been hypothesized as a factor involved in the onset of dry eye.7 Endocrine regulation of the lacrimal gland8 is apparent from its sexually dimorphic morphology and function: women experience dry eye problems, and especially Sjögren’s dry eye, more frequently than men. In male rodents, the glands are larger, contain larger acini, and show lower acinar cell density.9 Functionally, male glands secrete higher levels of IgA and secretory component.5,10,11 Castration of males results in the loss of sexual dimorphism. Glands assume a more female morphology, and the levels of IgA and secretory component are reduced. Treatment of castrated animals with androgens reestablishes male morphology and increases IgA and secretory component output.1,12–15 Of note, in mouse models of autoimmune disease, androgen treatment can suppress the immunopathologic lesions of the lacrimal glands.16 –18 Topical androgen application has been suggested for treatment of both Sjögren’s and non-Sjögren’s dry eye syndrome.19 Androgens do not act alone on the lacrimal gland. Mice without androgen receptors do not have a deficit in lacrimation.19 Hypophysectomy or pituitary transplant can prevent androgen-induced restoration of tear volume, IgA, and secretory component levels after castration.1,12–15 The identity of this pituitary factor is unknown, but because transplanted pituitaries secrete high levels of prolactin, and prolactin treat23 24 McClellan et al. ment of dwarf mice has trophic effects on the lacrimal gland,8 it has been hypothesized to be a second hormone influencing lacrimal morphology and function. Short-term treatment with prolactin can restore the lacrimal expression of a number of genes that are altered after hypophysectomy, and can prevent dihydrotestosterone restoration of the levels of other genes, suggesting that prolactin modulates the lacrimal gland, both alone and in combination with androgens.1 These findings suggest that physiological levels of prolactin are required for the trophic actions of androgens on the lacrimal gland and that both hyper- and hypoprolactinemia may prevent this action. This hypothesis is consistent with the hormonal states in which dry eye is most common, but there is no convincing evidence in its favor. Because androgen receptors20 and prolactin and its receptor21 are expressed by the acinar cells of the lacrimal gland, this interaction may occur directly within the lacrimal gland. We have used two genetic mouse models of prolactin action; a transgenic mouse (PRLtg), hyperprolactinemic because of overexpression of rat prolactin,22 and a prolactin receptor knockout mouse (PRLR⫺/⫺) that has no prolactin receptors,23 to examine the hypothesis that prolactin is the pituitary factor involved in the function and maintenance of the lacrimal gland. These models allow two fundamental questions to be investigated using animals exposed to altered prolactin function from the early embryonic stage: Is prolactin essential for normal lacrimal development and function? Can increased levels of prolactin modulate lacrimal development? MATERIALS AND METHODS Mice The prolactin receptor knockout mouse (PRLR⫺/⫺) was generated by replacement of exon 5 of the PRLR gene, which encodes cysteine residues essential for ligand binding and receptor activation, with the NEO cassette.23 PRLR⫺/⫺ mice used in these experiments were derived from chimeric animals made using E14 embryonic stem cells (129/OlaHsd) bred to 129/Sv Pas mice. Genetically similar wild-type control mice (PRLR⫹/⫹ ) were obtained from heterozygous matings. PRLR⫺/⫺ animals show defects in fertility,23,24 bone development,25 mammary gland development,23,24,26,27 and maternal behavior,28 but a normal immune system29 and prostate development (Ormandy et al., manuscript in preparation). The prolactin transgenic (PRLtg) animal22 was generated by microinjection of a plasmid construct driving expression of the rat prolactin gene by the metallothionein promoter into C57Bl/6xCBA-f2 embryos, resulting in constitutive and generalized tissue expression. Genetically similar wild-type control animals (PRLwt), were generated from heterozygous matings. This animal shows altered mammary development and tumors30 and prostate enlargement and hyperplasia.22 All mice were housed in a 12-hour day/ night cycle at 22°C and 80% relative humidity with access to food and water ad libitum and were treated according to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Histology and Morphometric Analysis Individual lacrimal and Harderian glands were excised, laid flat onto filter paper (4 M; Whatman, Clifton, NJ) to maintain morphology during fixation, and fixed overnight in 10% neutral buffered formalin. Specimens were paraffin embedded, sectioned at 5 ␮m, and stained with hematoxylin-eosin. Specimens were photographed using a microscope (DMRB; Leica, Heidelberg, Germany) fitted with a CCD video camera (model 3; Sony, Tokyo, Japan) coupled to an image analysis program (Q500MC; Leica) running on a desktop computer. Acinar areas were measured using images captured at ⫻10 magnification. Epithelial cells were counted in images captured at ⫻20 magnification using the image analysis software. IOVS, January 2001, Vol. 42, No. 1 2-D Gel Analysis of Tear Proteins Mouse tears were collected from the eye with a 3-mm2 piece of filter paper by insertion between the orbit and lower eyelid of anesthetized mice. The portion of tear-soaked paper was added to 125 ␮l of a solution containing 5 M urea, 2 M thiourea, 100 mM dithiothreitol, 2% 3-([3-cholamidopropyl]dimethylammonio)-2-hydroxy-1-propanesulfonate (CHAPS), 2% sulfobetaine 3-10, 0.5% 3/10 Pharmalytes (Amersham Pharmacia Biotech, Amersham, UK), 40 mM Tris, and 0.001% bromophenol blue for 1 hour, after which the solution was vortexed. Sevencentimeter pH 3-10 immobilised pH gradient strips (IPGs; Amersham Pharmacia Biotech) were rehydrated with solution for 6 hours. Isoelectric focusing was conducted using a 2-dimensional gel electrophoresis apparatus (Multiphor II; Amersham Pharmacia Biotech) at 20°C and was maintained for 14,000 Vh. Two-dimensional (2-D) separation was conducted using 12.5% polyacrylamide gels in which the immobilized pH gradient strips were embedded with 1% (wt/vol) agarose and then run at 10 mA/gel. The gels were fixed and silver stained, and protein maps were constructed. Enzyme-Linked Immunosorbent Assay To determine IgA levels in tears, an isotype-specific sandwich enzymelinked immunosorbent assay (ELISA) was performed, using a direct plate binding assay. Plates were coated with a goat anti-mouse IgA capture antibody (PharMingen, San Diego, CA), and bound immunoglobulin was revealed by a second biotinylated goat anti-mouse IgA (Southern Biotechnology, Birmingham, AL) followed by streptavidinalkaline phosphatase and specific substrate (p-nitrophenylphosphate [pNPP]). Plates were read using an ELISA plate reader at 405 nm. Quantification was performed by comparison with a standard curve established using purified mouse IgA (PharMingen). Western Blot Analysis The sample was separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE; 8% acrylamide), transferred to membranes (Immobilon P; Amersham) using a semidry transfer apparatus (Bio-Rad; Herts, UK) and probed with either a polyclonal rabbit antiserum against rat secretory component, cross-reacting with the murine secretory component (courtesy of Jean Paul Vaerman), or a biotinconjugated rabbit polyclonal anti-murine IgA (Zymed Laboratories, South San Francisco, CA). For the IgA blots, membranes were incubated directly with streptavidin-horse radish peroxidase (Amersham), and revealed using enhanced chemiluminescence (ECL; Amersham). For the secretory component blots, membranes were incubated with a biotin-conjugated anti-rabbit antibody (Vector, Peterborough, UK), before streptavidin-horse radish peroxidase and ECL assays. All incubations and washes were in phosphate-buffered saline (PBS) with 0.05% Tween. Statistical Analysis Cell density, acinar area, and porphyrin secretions of lacrimal and Harderian glands were compared using an unpaired, two-tailed Students t-test (Statview 4.0; Abacus, Berkeley, CA). Incidence of conjunctivitis was analyzed using Kaplan–Meier survival analysis (Statview 4.0) and probabilities were calculated using the Mantle–Cox log rank method. RESULTS Lacrimal Gland Development The role of prolactin in the development of the sexually dimorphic morphology of the lacrimal glands was examined in mature animals (12–16 weeks) by comparison of PRLR⫺/⫺ with PRLR⫹/⫹ and PRLtg with PRLwt. The lacrimal is a tubuloalveolar gland composed of ducts (acini) without a clearly distinguished lumen. The ducts are surrounded by a fibrous IOVS, January 2001, Vol. 42, No. 1 Prolactin and Lacrimal Development 25 FIGURE 1. Lacrimal histology from mature animals (12–16 weeks of age), stained with hematoxylin-eosin: (A, B, and C) females; (D, E, and F) males. (A, D) Histology in PRLR⫹/⫹ animals was very similar to PRLwt animals (not shown). (B, E) PRLR⫺/⫺ animals and (C, F) PRLtg animals. Original magnification, ⫻20. basement membrane and contain two major cell types: the secretory epithelial cells (acinar cells), identified by large round nuclei located at the basement membrane surface of a cytoplasm replete with secretory vesicles, and less frequent myoepithelial cells closely associated with the basement membrane and displaying elongated nuclei. A mainly acellular stroma fills the intraductal space. These glands showed typical sexual dimorphism; female glands (Figs. 1A, 1B, 1C) showed smaller acini and increased acinar cell density when compared with male glands (Figs. 1D, 1E, 1F). The degree of sexual dimorphism was quantified by measurement of lacrimal acinar area using image analysis software, and acinar and acinar cell density by direct counting per field. Results were expressed as a percentage of female control levels, using elements per field, which correlated very well with direct area measurements. This allows comparison between the different mouse strains used. The sexual dimorphism of the lacrimal gland was easily detected by this technique. Male control animals (PRLR⫹/⫹, PRLwt) had approximately 75% of the female control number of acini per field because of an increase in average acinar area. Acinar cell density decreased to 75% of female levels (Fig. 2). In PRLtg females, hyperprolactinemia caused the lacrimal glands to assume a hyperfemale state with average acini per field increasing to 130% (P ⫽ 0.0016) of control because of a decline in average acinar area associated with a similar increase (25%, P ⫽ 0.0087) in acinar cells per field. The morphology of male PRLtg glands remained unchanged (P ⫽ 0.07). These changes increased the degree of sexual dimorphism seen in these animals. In female PRLR⫺/⫺ animals, loss of the prolactin receptor had no effect on the morphology of the lacrimal gland. In PRLR⫺/⫺ males there was a small (5%) but significant (P ⬍ 0.0001) increase in acini per field, caused by a decrease in acinar area associated with an 11% (P ⫽ 0.0034) increase in acinar cell density. Thus, in this model we also saw the acinar area decline and cell density increase, but in males not females, indicating a decrease in the degree of sexual dimorphism (Fig. 2). The small magnitude of this effect may not be physiologically relevant. To determine whether we had missed a transient effect of hypoprolactinemia in the PRLR⫺/⫺ animals during the onset of sexual dimorphism at puberty, we examined the lacrimal glands of animals at 4, 6, and 8 weeks of age (Fig. 3). The onset of sexual dimorphism was seen in females as a slight decrease in acinar area and slight increase in acinar cell number and in males as an increase in acinar area and dramatic decrease in acinar cell number, indicating that it is the male gland that most alters its morphology during puberty, consistent with the hypothesized trophic role of androgens. There was no difference in the rate of onset of sexual dimorphism between PRLR⫹/⫹ and PRLR⫺/⫺ animals of either gender. FIGURE 2. Lacrimal morphometric analysis from mature animals. The number of acini and acinar cells were counted in 10 random microscope fields per animal, three to five animals per group, and statistically analyzed. Results are expressed as a percentage of female PRLR⫹/⫹ for the knockout model and as a percentage of female PRLwt values for the transgenic model. 26 McClellan et al. IOVS, January 2001, Vol. 42, No. 1 fewer acinar cells than females. Quantification, as used for the lacrimal glands, showed smaller but similar effects of hyper- or hypoprolactinemia on acinar size or cell number in adults (Fig. 6). Female PRLtg glands showed an 8% decrease in acinar area that failed to reach statistical significance (P ⫽ 0.15) and a 10% (P ⫽ 0.05) increase in acinar cell density. Male PRLtg animals showed no changes. Female PRLR⫺/⫺ morphology remained unchanged, but male glands showed a 2% decrease in area that failed to reach statistical significance (P ⫽ 0.09) and a 10% (P ⫽ 0.007) increase in cell density. These changes exactly mirror those seen in the lacrimal gland, but their small magnitude places them close to the level of detection for the quantification technique that was used. Female acini, and male acini to a lesser extent, contained solid accretions of porphyrin in the 129 mouse strain used to make the prolactin receptor knockout (Figs. 5A 5D). Hypo- FIGURE 3. Lacrimal morphometric analysis during puberty. Lacrimal histology was analyzed in 10 random microscope fields per animal, three to five animals per group, during early puberty (4 weeks), midpuberty (6 weeks), late puberty (8 weeks), and at maturity (12 weeks). Results are expressed as the number of acinar cells or acini in PRLR⫹/⫹ or PRLR⫺/⫺. We examined the role of prolactin in the maintenance of sexual dimorphism by castrating PRLR⫹/⫹ and PRLR⫺/⫺ males and examining lacrimal morphology 21 days later (Fig. 4). In PRLR⫹/⫹ animals, castration resulted in a 10% decrease in acinar area and a 12% decrease in acinar cell density. PRLR⫺/⫺ animals underwent much the same change in acinar area, resulting in a 7% (P ⬍ 0.0001) difference. Acinar cell density showed a greater proportional change, so that no significant difference (P ⫽ 0.43) between genotypes was then seen in acinar cell density. These results indicate that the small reduction in acinar area resulting from a loss of the prolactin receptor is independent of androgen action, but that acinar cell number may be influenced by an interaction between prolactin and androgen. Again, however, the small magnitude of this effect calls into question its physiological relevance. Harderian Gland Development Histologic investigation (Fig. 5) showed that the Harderian glands of the mouse strains used in this study had larger acini with a defined lumen and more acinar cells than the lacrimal glands. They also exhibited sexually dimorphic characteristics similar to the lacrimal glands. Male glands had larger acini and FIGURE 4. Effect of castration on male PRLR⫹/⫹ and PRLR⫺/⫺ lacrimal glands. Mature males were castrated, and lacrimal morphology was analyzed 21 days later in 10 random microscope fields per animal, three to five animals per group. IOVS, January 2001, Vol. 42, No. 1 Prolactin and Lacrimal Development 27 FIGURE 5. Harderian gland histology from mature animals (12–16 weeks of age) stained with hematoxylin-eosin: (A, B, and C) females; (D, E, and F) males. Panel descriptions are the same as in Figure 1. Porphyrin accretions are seen in the acini as dark-staining oval-shaped figures. Original magnification, ⫻20. prolactinemia resulted in a complete loss of these solid porphyrin accretions in both males and females (Figs. 4B, 4E), quantified by counting accretions per field (Fig. 7). In males a 97% loss was seen (P ⫽ 0.0026), and in females an 83% (P ⬍ 0.0001) reduction occurred. The mouse strain used for construction of the transgenic model shows virtually no porphyrin accretions, making this analysis difficult. In males a 33% (P ⫽ 0.096) increase was seen and in females a 266% (P ⫽ 0.16) increase was seen, but the overall small number of accretions found prevented reliable statistical analysis of this effect. FIGURE 6. Harderian gland morphometric analysis from mature animals. The number of acini and acinar cells were counted in 10 random microscope fields per animal, three to five animals per group, and statistically analyzed. Results are expressed as a percentage of female PRLR⫹/⫹ counts for the knockout model and as a percentage of female PRLwt counts for the transgenic model. Lacrimal and Harderian Function To determine whether hypoprolactinemia alters the function of the lacrimal or Harderian glands, we took a number of approaches. Tear proteins were analyzed by 2-D gels, IgA and secretory component levels were analyzed by ELISA and Western blot, and alteration to the genetic susceptibility of the 129 mouse strain to conjunctivitis was searched for during an aging study. Silver staining of 2-D gels of tears from PRLR⫹/⫹ and PRLR⫺/⫺ animals and comparison to consensus gels of mouse serum identified many spots specific to the tear film. Repeated tear sampling and 2-D analysis (10 replicates) indicated that FIGURE 7. Harderian gland porphyrin accretions. Accretions were counted per ⫻10 field, 10 random fields per animal, three to five animals per group. A very low frequency of porphyrin accretions in the mouse strain used to produce the PRLwt and PRLtg animals prevented an accurate analysis of the role of hyperprolactinemia. 28 McClellan et al. IOVS, January 2001, Vol. 42, No. 1 none of these spots showed reproducibly altered patterns of expression (data not shown), leading to the conclusion that synthesis and secretion of the major tear proteins was unchanged. Individual spots could, however, show different relative levels in a single experiment, underlining the need for multiple replicates for reliable results using this technique. The secretory immune function of the eye is maintained through the concentration of IgA in tears. IgA is synthesized by the acinar cells of the lacrimal gland and is transferred to tears in association with the polymeric IgA receptor, secretory component. IgAs protect the cornea and conjunctiva from inflammatory and infectious disease. To determine whether the levels of IgA and secretory component in tears is dependent on prolactin, we measured these species by Western blot and ELISA (Fig. 8). In tears from PRLR⫺/⫺ animals, Western blot analysis showed IgA and secretory component levels were identical with levels found in tears from PRLR⫹/⫹ animals. This was confirmed by ELISA for IgA. These experiments discount any essential role for prolactin in the maintenance of the major molecular species involved in the secretory immune system of the eye. We used a genetic characteristic of the 129 strain to test for reduced ocular immune function. The 129 mouse strain is susceptible to conjunctivitis, which begins as mild suppurative palpebral conjunctivitis and progresses to the mucocutaneous junction at the exit of the meibomian duct, where a suppurative process develops within and adjacent to the duct, associated with the formation of small ulcers over the conjunctiva. A genetic deficiency in secretory immunity appears to be the cause.31 As the disease advances, the eyelid becomes swollen and the surrounding hair becomes matted, and in our facility, mice at this stage are culled. To determine whether the secretory immune system of the eye was further compromised in PRLR⫺/⫺ mice, we aged a group of PRLR⫹/⫹ and PRLR⫺/⫺ animals and compared their survival. Animals that died or were culled for other causes were removed from the study. By 18 months of age the overall culling rate due to conjunctivitis was, for females, PRLR⫹/⫹ 4 of 10 (40%) and PRLR⫺/⫺ 9 of 25 (36%), and for males, PRLR⫹/⫹ 8 of 18 (44%) and PRLR⫺/⫺ 3 of 14 (21%). These data were analyzed further by Kaplan–Meier survival analysis (Fig. 9). Calculation of probabilities indicated no significant difference between genotypes when analyzed without reference to gender (P ⫽ 0.16 Mantle–Cox log rank) or when analyzed separately by gender (males P ⫽ 0.82, females P ⫽ 0.98, Mantle–Cox log rank). These experiments detected no essential role for prolactin in the function of the secretory immune system of the eyes of animals predisposed to conjunctivitis. DISCUSSION The results demonstrate that in females, prolactin plays no essential role in the development or maintenance of the morphology or function of the lacrimal glands, but that prolactin in excess can alter lacrimal gland morphology. Application of these findings to the previous investigation of the endocrine control of the lacrimal gland revealed unseen difficulties in the interpretation of results. In experiments using hypophysectomy in females or female pituitary dwarf mice, it is now clear that loss of prolactin was not the cause of the alterations in lacrimal gland morphology and function. Prolactin is not the pituitary factor lost in these models that influences the lacrimal gland, nor is it the pituitary factor that modulates androgen action in these experimental paradigms. These experiments also show that the effects in females of pituitary transplant or prolactin injection were due to the hyperprolactinemic result of these manipulations and did not reveal an effect of normal FIGURE 8. Secretory component and IgA levels in the tears of PRLR⫹/⫹ and PRLR⫺/⫺ mice. (A) Western blot of IgA or secretory component in tears from PRLR⫺/⫺ (⫺/⫺) or PRLR⫹/⫹ (⫹/⫹) animals. Tear samples (10 ␮l) were separated by SDS-PAGE and transferred to membranes. (B) ELISA. Tear concentration of IgA. prolactin levels on the lacrimal glands. Thus, when prolactin treatment or pituitary transplant are combined with hypophysectomy, the effects of hyperprolactinemia are overlaid on the independent effects of hypophysectomy. Effects previously attributed to a physiological role of prolactin are in fact due to superphysiological levels of the hormone, which causes female lacrimal glands to assume a hyperfemale morphology. Translating this result to humans suggests hyperprolactinemia may Prolactin and Lacrimal Development IOVS, January 2001, Vol. 42, No. 1 FIGURE 9. Kaplan–Meier survival analysis of conjunctivitis in PRLR⫹/⫹ and PRLR⫺/⫺ mice. The 129 strain used to make the PRLR⫹/⫹ and PRLR⫺/⫺ animals carries a genetic susceptibility to suppurative conjunctivitis. Survival analysis showed no difference in susceptibility between PRLR⫹/⫹ (closed symbols) and PRLR⫺/⫺ animals (open symbols) when analyzed without reference to gender (top) or by gender (bottom). predispose to dry eye, and support for this conclusion has been published. A correlative study7 of serum hormone levels and parameters of lacrimal gland activity, such as tear osmolarity, volume flow, and turnover, found that patients in menopause not using hormone replacement therapy had a positive correlation between tear volume and testosterone, but in women using hormone replacement therapy a strong negative correlation was found between serum prolactin level and multiple parameters of lacrimal function. In males, a small but significant requirement for prolactin in the establishment of sexual dimorphism was found, with both acinar area and cell number affected. This effect was very small, however, and may not affect the physiology of the gland. After castration, the difference between genotypes on acinar cell number was lost, indicating an interaction between prolactin and androgen to control acinar cell density. A similar situation occurs in the androgen-regulated ventral prostate and seminal vesicle. Both glands are lighter in prolactin knockout animals.32 The external adnexa of the eyes of mice (and all other species with a third eyelid) also includes the Harderian gland, located within the orbit behind the eye and almost encircling the optic nerve. It is found in humans in vestigial form during embryonic development and occasionally as a developmental abnormality.33 This gland adds lipids to the tear film through a duct that opens onto the surface of the nictitating membrane. 29 The gland also contains porphyrins, which are thought to be involved in sensing day length.34 Neonatal rat pups with undeveloped eyes or blind moles with vestigial eyes continue to respond to changed photoperiod when their eyes are removed, but not when their eyes and Harderian glands are removed.33,35 A number of these photoperiod responses involve the pineal gland, and the Harderian gland synthesizes melatonin and contains melatonin receptors, suggesting that it may have endocrine activity.33 The Harderian gland is sexually dimorphic and sensitive to steroid and pituitary hormones including prolactin.36 Our results indicate that the Harderian gland responds to prolactin in the same way as the lacrimal gland but that it is less sensitive, resulting in effects at the level of detection of our techniques. An essential role for prolactin was found in the formation of porphyrin accretions by the Harderian glands of male and female mice. Because testosterone levels in male PRLR⫺/⫺ animals are normal25 this observation establishes prolactin as a major and essential hormone controlling porphyrin accumulation in mice, as hypothesized from hypophysectomy and prolactin-bromocriptine treatment studies in rodents.36 Why prolactin should control accumulation of porphyrins in the Harderian gland remains an open question, but given prolactin’s diverse reproductive actions23,24 and the photo period sensing and signaling ability of the Harderian gland, it is tempting to speculate that it may have a role in the control of seasonal breeding. It is important to distinguish between the endocrine state produced by an absence of prolactin action and that produced by hyperprolactinemia. These conditions can be considered to be separate endocrine states and demonstrate that the failure to show an essential role for a hormone in a process does not indicate that an excess of that hormone will similarly be without effect. This is the case with prolactin in the female lacrimal gland. Although not essential for normal development, hyperprolactinemia produces a hyperfemale morphology that may predispose to dry eye. References 1. Azzarolo AM, Bjerrum K, Maves CA, et al. Hypophysectomy-induced regression of female rat lacrimal glands: partial restoration and maintenance by dihydrotestosterone and prolactin. Invest Ophthalmol Vis Sci. 1995;36:216 –226. 2. Serrander A, Peek K. Changes in contact lens comfort related to the menstrual cycle and menopause: a review of articles. J Am Optom Assoc. 1993;64:162–166. 3. Stern M, Beuerman R, Fox R, et al. The pathology of dry eye: the interaction between the ocular surface and lacrimal glands. Cornea. 1998;17:584 –589. 4. Lemp M. Report of the National Eye Institute/Industry Workshop on clinical trials in dry eyes. CLAO. 1995;21:221–232. 5. Sullivan D, Bloch K, Allansmith M. Hormonal influence on the secretory immune system of the eye: androgen control of the secretory component production by the rat exorbital gland. Immunology. 1984;52:239 –246. 6. Fox R, Tornwall J, Michelson P. Current issues in the diagnosis and treatment of Sjögren’s syndrome. Curr Opin Rheumatol. 1999;11: 364 –371. 7. Mathers W, Stovall D, Lane J, Zimmerman M, Johnson S. Menopause and tear function: the influence of prolactin and sex hormones on human tear production. Cornea. 1998;17:353–358. 8. Sullivan D, Wickham L, Rocha E, et al. Influence of gender, sex steroid hormones, and the hypothalamic-pituitary axis on the structure and function of the lacrimal gland. In: Sullivan D, et al., eds. Lacrimal Gland, Tear Film, and Dry Eye Syndromes 2. New York: Plenum Press; 1998:11– 42. 9. Cornell–Bell A, Sullivan D, Allansmith M. Gender-related differences in the morphology of the lacrimal gland. Invest Ophthalmol Vis Sci. 1985;26:1170 –1175. 30 McClellan et al. 10. Sullivan D, Bloch K, Allansmith M. Hormonal influence on the secretory immune system of the eye: androgen regulation of secretory component levels in rat tears. J Immunol. 1984;132:1130 – 1135. 11. Sullivan D, Allansmith M. Hormonal regulation of the secretory immune system in the eye: androgen modulation of IgA levels in tears of rats. J Immunol. 1985;134:2978 –2982. 12. Sullivan D, Allansmith M. Hormonal modulation of tear volume in the rat. Exp Eye Res. 1986;42:131–139. 13. Sullivan D, Block L, Pena J. Influence of androgens and pituitary hormones on the structural profile and secretory activity of the lacrimal gland. Acta Ophthalmol Scand. 1996;74:421– 435. 14. Gao J, Lambert RW, Wickham LA, Banting G, Sullivan DA. Androgen control of secretory component mRNA levels in the rat lacrimal gland. J Steroid Biochem Mol Biol. 1995;52:239 –249. 15. Sullivan D, Allansmith M. Hormonal influence on the secretory immune system of the eye: endocrine interactions in the control of IgA and secretory component levels in tears of rats. Immunology. 1987;60:337–343. 16. Sullivan DA, Ariga H, Vendramini AC, et al. Androgen-induced suppression of autoimmune disease in lacrimal glands of mouse models of Sjögren’s syndrome. Adv Exp Med Biol. 1994;350:683– 690. 17. Sullivan DA, Rocha FJ, Sato EH. Influence of androgen therapy on lacrimal autoimmune disease in a mouse model of Sjögren’s syndrome. Adv Exp Med Biol. 1995;371B:1199 –1202. 18. Sullivan D, Edwards J. Androgen stimulation of lacrimal gland function in mouse models of Sjögren’s syndrome. J Steroid Biochem Mol Biol. 1997;60:237–245. 19. Sullivan D, Krenzer K, Sullivan B, et al. Does androgen insufficiency cause lacrimal gland inflammation and aqueous tear deficiency? Invest Ophthalmol Vis Sci. 1999;40:1261–1265. 20. Sullivan DA, Edwards JA, Wickham LA, et al. Identification and endocrine control of sex steroid binding sites in the lacrimal gland. Curr Eye Res. 1996;15:279 –291. 21. Mircheff AK, Warren DW, Wood RL, Tortoriello PJ, Kaswan RL. Prolactin localization, binding, and effects on peroxidase release in rat exorbital lacrimal gland. Invest Ophthalmol Vis Sci. 1992;33: 641– 650. 22. Wennbo H, Kindblom J, Isaksson O, Tornell J. Transgenic mice overexpressing the prolactin gene develop dramatic enlargement of the prostate gland. Endocrinology. 1997;138:4410 – 4415. 23. Ormandy CJ, Camus A, Barra J, et al. Null mutation of the prolactin receptor gene produces multiple reproductive defects in the mouse. Genes Dev. 1997;11:167–178. View publication stats IOVS, January 2001, Vol. 42, No. 1 24. Binart N, Helloco C, Ormandy C, et al. Rescue of pre-implantatory egg development and embryo implantation in prolactin receptor deficient mice after progesterone administration. Endocrinology. 2000;141:2691–2697. 25. Clement–Lacroix P, Ormandy CJ, Lepescheux L, et al. Osteoblasts are a new target for prolactin: analysis of bone formation in prolactin receptor knockout mice. Endocrinology. 1999;140:96 – 105. 26. Ormandy CJ, Binart N, Kelly PA. Mammary gland development in prolactin receptor knockout mice. J Mamm Gland Biol Neoplasia. 1997;2:355–364. 27. Ormandy C, Horseman N, Naylor M, et al. Mammary gland development. In: Horseman N, ed. Prolactin. Amsterdam: Kluwer; In press. 28. Lucas BK, Ormandy CJ, Binart N, Bridges RS, Kelly PA. Null mutation of the prolactin receptor gene produces a defect in maternal behaviour. Endocrinology. 1998;139:4102– 4106. 29. Bouchard B, Ormandy C, Di Santo JP, Kelly PA. Immune system development and function in prolactin receptor deficient mice. J Immunol. 1999;163:576 –582. 30. Wennbo H, Gebre–Medhin M, Gritli–Linde A, et al. Activation of the prolactin receptor but not the growth hormone receptor is important for induction of mammary tumors in transgenic mice. J Clin Invest. 1997:2744 –2751. 31. Sundberg J, Brown K, Bates R, Cunliffe–Beamer T, Bedigian H. Suppurative conjunctivitis and ulcerative blepharitis in 129/J mice. Lab Anim Sci. 1991;41:516 –518. 32. Steger R, Chandrashekar V, Zhao W, Bartke A, Horseman N. Neuroendocrine and reproductive functions in male mice with targeted disruption of the prolactin gene. Endocrinology. 1998;139: 3691–3695. 33. Chieffi G, Baccari GC, Di Matteo L, et al. Cell biology of the Harderian gland. Int Rev Cytol. 1996;168:1– 80. 34. Wetterberg L, Geller E, Yuwiler A. Harderian gland: an extraretinal photoreceptor influencing the pineal gland in neonatal rats? Science. 1970;167:884 – 885. 35. Pévet P, Heth G, Hiam A, Nevo E. Photoperiod perception in the blind male rat (Sphalex chrenbergi Mehring): involvement of the Harderian glands, atrophied eyes and melatonin. J Exp Zool. 1984; 232:41–50. 36. Payne AP, Shah SW, Marr FA, et al. Hormones and the control of porphyrin biosynthesis and structure in the hamster harderian gland. Microsc Res Tech. 1996;34:123–132.