| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1From the Departments of Cell and Neurobiology, 3Physiology and Biophysics, and the 2Doheny Eye Institute, Keck School of Medicine, University of Southern California, Los Angeles, California.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
METHODS. Corneas were injured with scalpel blades, and, 2 weeks later, LGs were collected for immunocytochemistry and Western blot analysis. Lacrimal fluid was collected under basal- and pilocarpine-stimulated conditions for protein determination and Western blot analyses.
RESULTS. There were significant increases of immunoreactivity for prolactin, TGF-ß1, and EGF in duct cells during pregnancy and after CI, most prominent in pregnant animals with CI. Pregnancy decreased baseline lacrimal fluid secretion, whereas CI did not have a noticeable effect; pregnancy and CI combined resulted in increased fluid production. Pregnancy and CI each increased pilocarpine-induced lacrimal fluid production, whereas protein concentrations were decreased. Prolactin, TGF-ß1, and EGF were detected in LG by Western blot analysis but were minimally detectable in lacrimal fluid. RTLA+ and CD18+ cells were redistributed from periductal to interacinar sites during pregnancy and after CI, most prominent in pregnant animals with CI.
CONCLUSIONS. Like pregnancy, CI is associated with redistribution of immune cells from periductal to interacinar sites and enhanced immunoreactivity of prolactin, TGF-ß1, and EGF in ductal cells. Although baseline lacrimal fluid secretion varied, the glands of all three experimental groups produced significant amounts of fluid in response to pilocarpine, but protein concentrations were decreased.
Prolactin and several other paracrine factors have been documented to be present in the lacrimal glands, and it has been proposed that they play a role in regulating lacrimal secretion.2 3 4 5 6 7 Growth factors have also been detected in the lacrimal fluid,2 3 8 9 10 although, apart from supporting corneal wound healing, their roles in ocular surface function are still unclear.
Much previous work has suggested that sex hormones, prolactin, and growth factors all contribute to changes in the lacrimal gland and the onset and progression of dry eye.11 12 13 Because there is a significant change of the hormonal milieu during pregnancy, pregnancy would seem to serve as an excellent model for studying the influences of normal, physiological endocrine changes on lacrimal function. In our earlier studies, we found that as pregnancy progresses in New Zealand White rabbits, immunolabeling of prolactin, basic fibroblast growth factor (FGF)-2, epidermal growth factor (EGF), and transforming growth factor (TGF)-ß1, becomes more intense within intra- and interlobular ducts. We also found that the number of immune cells in periductal foci decrease, whereas the number of immune cells at interacinar sites in the gland increase, suggesting that these cells are drawn away from the periductal foci and toward the interacinar interstitium.14 15
Acute changes in lacrimal fluid secretion rates and longer-term changes in gene expression occur within the lacrimal gland in response to corneal injury (CI).15 16 17 The changes in growth factor expression are accompanied by immunoarchitectural changes reminiscent of those that occur during pregnancy. Because the occurrence of such changes in pregnancy seemed to us consistent with an overall heightened state of immune readiness,14 we hypothesized that (1) pregnancy would enhance the changes of immunoarchitecture and growth factor expression induced by CI, (2) lacrimal secretory function would change during both pregnancy and after CI, and (3) pregnancy would exacerbate the functional changes associated with CI. The results reveal a more complex pattern of interactions between the influences of pregnancy and corneal insult.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Nonpregnant animals without CI were considered the control group. The experimental groups were: pregnant without CI, nonpregnant with CI, and pregnant with CI.
CI was performed on the 15th day of pregnancy. A scalpel blade was used to penetrate the corneal stroma in a five-point dice pattern, whereas the animals were under general anesthesia. The procedure has been detailed previously.18 Pregnant rabbits were killed on the 29th day of pregnancy, and rabbits in the nonpregnant with CI group were killed 14 days after CI.
Inferior lacrimal glands were removed, and some tissue fragments were placed in optimal cutting temperature (OCT) compound (Miles, Elkhart, IN) and rapidly frozen with liquid nitrogen. Some fragments were fixed and processed for epoxy embedding for light and electron microscopy, as previously described.19 Additional fragments were used for determination of protein concentration and for Western blot analysis. All studies conformed to the standards and procedures for the proper care and use of animals, as described in the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.
Animals used in these studies were: 5 nonpregnant without CI, 16 pregnant without CI, 6 nonpregnant with CI, and 11 pregnant with CI.
Immunohistochemistry
Tissue fragments were cryosectioned at 6 to 8 µm in thickness and affixed to polyfrost glass slides (Fisher Scientific, Hampton, NH). Sections were fixed briefly with acetone or 4% paraformaldehyde and processed using the standard immunohistochemistry protocol.
Protocols for immunohistochemical staining were run in parallel for sections from the control and experimental groups. Two primary antibodies were used to label immune cells: goat anti-rabbit T lymphocyte antigen (RTLA; Cedarlane, Ontario, Canada), and mouse anti-rabbit CD18 (Spring Valley, Woodbine, MD). RTLA marks thymus-derived immune cells in rabbits and CD18 labels bone-marrow–derived immune cells. Other antibodies used were: guinea pig anti-rabbit prolactin (obtained from Alfred F. Parlow, National Institute of Diabetes and Digestive and Kidney Diseases [NIDDK] Pituitary Hormones and Antisera Center), chicken anti-human TGF-ß1 (R&D Systems, Minneapolis, MN), and mouse anti-human EGF (CalBiochem, Cambridge, MA).
Secondary antibodies used for immunohistochemistry were goat anti-mouse IgG for EGF and CD18, donkey anti-goat IgG for RTLA, goat anti-guinea pig IgG for prolactin, and goat anti-chicken IgG for TGF-ß1. Slides were treated with ABC reagent (Vector Laboratories, Burlingame, CA) and diaminobenzidine and mounted (Immumount; Shandon/Lipshaw, Pittsburgh, PA). Negative controls specimens consisted of exposure to secondary antibody and ABC reagent without exposure to primary antibody. Sections of rabbit anterior pituitary gland served as the positive control for prolactin. Slides were counterstained with methylene blue.
Image Analysis
Immunohistochemical data for RTLA and CD18 were quantified using a computer-assisted, digital image-analysis system. The detailed procedure has been described previously.15
The objective of the image analysis was to determine whether redistribution of immune cells from periductal sites to interstitial sites occurs in all the experimental groups.
Lacrimal Gland Fluid Collection
Lacrimal ducts were cannulated with small-bore polyethylene tubing (inner diameter [ID]: 0.28 mm, outer diameter [OD]: 0.61 mm; BD Bioscience, Sparks, MD) that had been pretreated with protease inhibitor. Basal lacrimal secretion was collected for 15 minutes, followed by intravenous administration of pilocarpine (1 mL, 1.2 mg/mL) and three additional collections for 10 minutes each. Collecting tubes with the fluids were weighed and the volume determined from the weight differential. These volumes divided by the duration of the collection period gave the flow rate (in microliters/minute).
Protein Determination, SDS-PAGE, and Western Blot Analysis
Protein concentrations in lacrimal gland lysates and lacrimal fluid were determined according to the Bradford protein protocol on a microplate reader (SOFTmax software; E-max microplate reader; Molecular Devices, Sunnyvale, CA).
Samples for polyacrylamide gels (SDS-PAGE) were prepared so that the same amount of protein could be loaded into each lane on the gel. Polyacrylamide gels at 12% were run according to standard protocols.
Protein loads were 100 µg per lane for analyses of TGF-ß1 and EGF and 50 µg per lane for analyses of prolactin. Proteins were transferred onto nitrocellulose membranes and exposed to primary antibody overnight at 4°C (1:1000, guinea pig anti-rabbit prolactin, obtained from Alfred F. Parlow, NIDDK; 1:500, mouse anti-human TGF-ß1; Chemicon Inc., Temecula, CA; 1:500, goat anti-human EGF; R&D Systems Inc.). Membranes were then exposed to the appropriate dye-conjugated secondary antibodies (IRDye 800; Rockland, Gilbertsville, PA): goat anti-guinea pig IgG (1:3000) for prolactin; goat anti-mouse IgG (1:2000) for TGF-ß1; and donkey anti-goat IgG (1:2000) for epidermal growth factor [EGF]. Reference samples were: sheep prolactin, 22.5 kDa (Sigma-Aldrich, St. Louis, MO), and human EGF, 6.35 kDa (Oncogene, Cambridge, MA). Near-infrared fluorophores (IRDye-800; Rockland) were determined (Odyssey Infrared Fluorescence Imaging System; LiCor Instruments, Lincoln, NB), and densitometry was conducted on computer (Labworks 4.0; Ultra-Violet Products, Inc., Upland, CA).
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
Transforming Growth Factor-ß1.
A major band was demonstrated at approximately 23 kDa in lysates of lacrimal gland from the control and all three experimental groups. Pregnancy increased its intensity by 10%, whereas CI decreased it by 80%. However, in pregnant rabbits with CI, there was virtually no change in intensity.
As for prolactin, TGF-ß1 was minimally detectable within lacrimal fluid. However, densitometric analysis demonstrated that the density in the three experimental groups was increased between 1.8- and 3.3-fold. The highest density was detected in nonpregnant animals with CI, which had the lowest density within the gland lysates.
Epidermal Growth Factor.
A major band at 
28 kDa, corresponding to the membrane-bound precursor of EGF, was detectable in lacrimal gland lysates from control and all experimental groups. It was minimally detectable in the lacrimal glands of nonpregnant animals without CI, and increased dramatically by either pregnancy or CI, with the strongest being observed in pregnant animals without CI, a 24-fold increase. However, the influences of pregnancy and CI appear to be antagonistic, suggested by the finding that the intensity of EGF in pregnant animals with CI being intermediate between pregnant without CI and nonpregnant with CI groups.
EGF was minimally detectable within the lacrimal fluid, with pregnancy and CI appearing to increase its intensity, a 4-fold increase for pregnancy and 1.5-fold increase for CI. However, its intensity was slightly lower when the effects of pregnancy and CI were combined (i.e., in comparison to nonpregnant animals without CI).
Immune Cell Redistribution
RTLA+ Staining.
RTLA+ cells were observed in the lacrimal glands of all four groups, in both interacinar and periductal areas (Table 2) .
|
Periductal Cells.
Pregnancy decreased the number of periductal RTLA+ cells (P < 0.05). Whereas CI also decreased periductal RTLA+ cells in nonpregnant animals, the effect was only half as much as pregnancy alone (P < 0.01). CI did not significantly alter the number of periductal RTLA+ cells in pregnant rabbits (P > 0.05).
CD18+ Staining.
CD18+ cells were observed in all four groups, both interacinar and periductal (Table 3) .
|
Periductal Cells.
Pregnancy alone decreased the number of periductal CD18+ cells (P < 0.01), with a similar effect observed during CI. However, the two conditions did not have any synergistic effect on the number of periductal CD18+ cells.
Lacrimal Secretory Functions
Fluid Production.
In control animals, lacrimal fluid was secreted at a basal rate of 0.35 µL/min (Table 4) . In pregnant rabbits, the secretion rate decreased to 0.21 µL/min (P < 0.05). CI had no significant effect on the basal fluid secretion rate in nonpregnant animals, but it increased the basal rate nearly twofold in pregnant animals (P < 0.05).
|
Moreover, CI partially reversed the effect of pregnancy on the pilocarpine-induced fluid secretion rate, so that the rates in pregnant animals with CI fell between those of pregnant without CI and nonpregnant with CI. Although no significant difference of stimulated lacrimal fluid secretion was observed between the two injured groups (P = 0.21), there were differences between the fluid secretion rates among all other groups (P < 0.05).
Protein Secretion.
Protein secretion in pilocarpine-stimulated lacrimal fluids from rabbits of each group is shown in Table 5 . Whereas pregnancy more than doubled the pilocarpine-induced lacrimal fluid flow rate, it decreased the concentration of protein in the secreted fluid by 38% (Table 5 ; ratio of 22.23:13.75). However, taking into consideration that there was an increased volume of lacrimal fluid in response to pilocarpine in pregnant animals without CI, the total amount of protein secreted increased by 35% (Table 5 ; ratio of 22.23:29.97).
|
Like the changes in fluid secretion, CI partially reversed the effect of pregnancy in increasing the amount of protein secretion; therefore, the combined effect of pregnancy and CI increased the amount of protein secreted by 19.4% (Table 5 ; ratio of 22.23:26.57).
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Despite the overall similarity between the influences of pregnancy and CI on lacrimal immunoarchitecture, there are certain differences between these two states. Pregnancy is associated with greater decrease in the number of periductal RTLA+ cells, and the influence of pregnancy appears to override the influence of CI. On the other hand, CI is associated with greater increase in the number of interacinar RTLA+ cells, and the influence of CI on interacinar RTLA+ cells appears to override the influence of pregnancy. Pregnancy and CI are associated with similar decreases in the number of periductal CD18+ cells, but CI is associated with a greater increase in the number of interacinar CD18+ cells. Moreover, the influences of the two states on the number of interacinar CD18+ cells appear to be additive.
Pregnancy is associated with a significant decrease in basal lacrimal fluid flow rate but a 118% increase in the rate the gland produces fluid in response to pilocarpine stimulation. CI is associated with no change in basal lacrimal fluid production and only a 42% increase of pilocarpine-induced fluid production, but in the setting of pregnancy, CI significantly increases the basal rate of fluid production. While pregnancy increases pilocarpine-induced protein secretion, CI decreases it. Therefore, the combined effects of pregnancy and CI appear to increase pilocarpine-induced protein secretion to a lesser degree than pregnancy alone, suggesting that CI partially reversed the effect of pregnancy.
Our immunohistochemistry data showed that either pregnancy or CI, and the combination of these two states, result in increased immunoreactivity of prolactin, TGF-ß1, and EGF within the ductal epithelial cells. The origin of these elevated amounts of peptides is unclear. In pregnant animals, as in nonpregnant ones, the overall staining for prolactin, TGF-ß1, and EGF was minimal in the acinar cells. However, because the number of lacrimal acinar cells far outnumber ductal epithelial cells, the cumulative release of these peptides may still be significant, despite the minimal amounts detected by immunocytochemistry. Although these three peptides were only minimally detectable in the lacrimal fluids within the detectability of our instrument and the time parameters of our studies, we cannot rule out the significance of their presence in the fluids and potential important influences on the ocular surface, in both physiological and pathologic conditions, such as CI. We plan continued assays in the future to further clarify these results. We also cannot distinguish between the alternatives that ductal epithelial cells themselves may produce these peptides, or alternatively may be reabsorbing and transporting these peptides from the acinar effluent within the ductal lumen. In either case, the ductal epithelial cells must be viewed as a key component in regulation of the primary lacrimal fluid.
The specific signaling mechanism(s) underlying the redistribution of immune cells from periductal foci to interacinar areas during pregnancy and after CI remains to be clarified, as well as the significance of this event in altering lacrimal function. It is known that the acinar effluent is modified as it passes across the ducts, with exchanges of water, electrolytes, and proteins ultimately resulting in the final fluid exiting the gland.20 21 22 It is our working hypothesis that there is interaction between the acinar and ductal elements within the gland (i.e., the ductal epithelial cells serve in monitoring and/or modulating the primary lacrimal fluid secreted from the acinar cells that traverses the duct system). Our studies suggest that prolactin, TGF-ß1, and EGF within the ductal epithelial cells are likely to function in signaling periductal immune cells (Schechter JE, unpublished data, 2000), and the occasional immunolocalization of these peptides within the basal cytoplasm of the ductal epithelial cells appears to support that hypothesis.
Among the signals most salient for the present discussion are prolactin and TGF-ß1, paracrine mediators expressed by the lacrimal epithelium. The information that has emerged from studies of other tissues of the mucosal immune system, including intestines and the mammary gland, suggest that prolactin and TGF-ß1 likely support maturation of IgA+ plasmablasts and survival of IgG+ plasmacytes, while providing autocrine support for expression of the polymeric immunoglobulin receptors (pIgR) that mediate transcytosis of dimeric IgA through the epithelium.23 24 25 26 Reciprocally, the immune effector and regulatory tissues secrete mediators that influence epithelial cell functions. Dimeric IgA itself is known to be an important signaling molecule in other systems. In Sjögren’s syndrome, agonistic autoantibodies to M3 muscarinic acetylcholine receptors are thought to induce a state of functional quiescence in the epithelium,27 whereas interleukin (IL)-1 and tumor necrosis factor (TNF)-
are thought to impair both secretomotor neurotransmission and epithelial responses.28 29
Finally, the body’s endocrine environment exerts direct and indirect influences on the effector tissues and likely on the control tissues as well. It is well established that androgens support expression of pIgR by epithelial cells in the lacrimal gland,30 31 as in the prostate.32 The results of the present study demonstrate that the hormonal milieu of pregnancy modulates expression of paracrine mediators by the lacrimal epithelium. Dissecting the interacting signaling relationships will require substantial effort, but several parameters seem clear. The prevailing endocrine signals in pregnancy are elevated levels of estrogen and progesterone, as well as an estrogen-induced increase of pituitary prolactin production. Also of importance, bioavailable testosterone levels decrease, as a consequence of estrogen-stimulated hepatic sex hormone binding globulin production. Our preliminary studies appear to confirm that estrogen and progesterone support the patterns of lacrimal immunoarchitecture and paracrine mediator expression seen in pregnancy (Ding C, unpublished data, 2004). Estrogen, progesterone, and prolactin interact to support pIgR expression and IgA+ plasmacyte homing and survival in the mammary gland.33 34 35 The diffuse, interacinar distribution of immune cells of the lacrimal gland during pregnancy is strongly reminiscent of the immunoarchitecture of the mammary gland before parturition and during lactation, so we infer that an analogous underlying physiological rationale may direct the pregnancy-associated changes we have documented in lacrimal gland immunoarchitecture.
Other workers have previously shown that lacrimal epithelial cells upregulate their expression of TGF-ß1 and EGF within minutes after CI.19 It seems likely, although not established definitively, that activation of sensory nerves in the cornea initiates, and autonomic secretomotor nerves mediate, this response.26 It has been presumed that lacrimal TGF-ß1 and EGF interact with local mediators to support wound healing,36 although the demonstration that EGF receptor expression also is upregulated suggests that local, autocrine, and paracrine responses also occur. Our data, as presented herein, demonstrated that rabbit lacrimal gland also increases its expression of prolactin after CI.
Whatever the signals—systemic hormonal changes or neural signals triggered by ocular surface injury that initiate upregulation of lacrimal epithelial prolactin, TGF-ß1, and EGF—it appears that once expressed, these paracrine mediators elicit a stereotypical response, with specific manifestations particular to the underlying physiological state. One of the specific manifestations this study shows to be associated with pregnancy is a decrease in basal lacrimal gland fluid secretion accompanied by an increased capacity to produce fluid in response to pilocarpine stimulation. It is possible that the decrease in basal fluid secretion results from a decrease in corneal sensitivity, rather than from any of the intrinsic signals associated with this physiological state. Then, it is reasonable to hypothesize that either decreased basal fluid production rates, or decreased concentrations of protein in fluid elicited by reflex signals from the ocular surface, are responsible for the ocular surface disease that we documented in our rabbit model37 and for the increased frequency of dry eye complaints among women who are pregnant.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Submitted for publication August 5, 2005; revised November 11 and December 16, 2005; accepted February 9, 2006.
Disclosure: C. Ding, None; N. Chang, None; Y.-C. Fong, None; Y. Wang, None; M.D. Trousdale, None; A.K. Mircheff, None; J.E. Schechter, None
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked "advertisement" in accordance with 18 U.S.C. 
1734 solely to indicate this fact.
Corresponding author: Chuanqing Ding, Department of Cell and Neurobiology, University of Southern California, Keck School of Medicine, 1333 San Pablo Street, BMT 304, Los Angeles, CA 90089-9112; cding{at}usc.edu.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. Wang, C. T. Chiu, T. Nakamura, A. M. Walker, B. Petridou, M. D. Trousdale, S. F. Hamm-Alvarez, J. E. Schechter, and A. K. Mircheff Elevated prolactin redirects secretory vesicle traffic in rabbit lacrimal acinar cells Am J Physiol Endocrinol Metab, April 1, 2007; 292(4): E1122 - E1134. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
