Within the PPs, the number of T cells dropped in the TPN groups, compared with chow-fed (experiment 1) and Nutrenfed (experiment 2) animals. B cells dropped in TPN-fed animals in both experiments, but reached statistical significance only in the first experiment.
Within the IE lymphocytes, there was a significant increase in the percentage of T cells compared with B cells in the IG-TPN and the IV-TPN groups (Table 7), but there were no significant differences in the absolute number of T cells (Table 8). This shift toward T cells was caused by a significant decrease in B cells with IG-TPN and IV-TPN in both experiments.
Within the LP, both B cells and T cells dropped significantly in the IG-TPN and the IV-TPN groups, compared with the Chow group without shifts in the B-cell/T-cell percentages. The Nutren group also had a significant drop in T-cell, but not B-cell, values.
There were no significant changes in MLN cell populations (data not shown).
There were no significant changes in MLN T-cell subsets in any group (data not shown), and the percentage of CD4sup + and CD8sup + cells in total cells obtained from the GALT were constant in all groups. The only exception was a significant increase in the percentage of LP CD8sup + cells in TPN animals, compared with the Chow group in experiment 1 and a drop in Nutren animals (Table 9) in experiment 2.
There were changes in the absolute number of CD4sup + and CD8sup + cells, however (Table 10). Within the PPs, CD4sup + cells dropped in both TPN groups in experiment 1, but reached significance only in IV-TPN mice while dropping significantly in IG-TPN mice (experiment 2), compared with Chow or Nutren animals. Within the IE lymphocytes, both CD4sup + and CD8sup + cells dropped with IV-TPN feeding, compared with chow (experiment 1) and with IG-TPN compared with Chow and Nutren animals (experiment 2). CD8sup + cells in the IE lymphocyte dropped in both TPN-fed groups, compared with the Chow group (experiment 1). Within the LP, CD4sup + cells dropped with IG-TPN (experiments 1 and 2), IV-TPN (experiment 1), and Nutren (experiment 2) animals, compared with chow, whereas CD8sup + cells decreased significantly in IG-TPN and Nutren groups, compared with the Chow group (experiment 2).
Overall, the CD4sup + /CD8sup + ratio (Table 11) remained constant in all areas except the LP, where IG-TPN and IV-TPN decreased significantly, compared with chow in experiment 1, and IG-TPN dropped significantly from both chow-fed and Nutren-fed animals in experiment 2.
There is substantial clinical [13-15] and experimental [16,17] evidence implicating route of nutrient administration as an important factor in reducing septic morbidity following injury, drawing attention to the gut as an important organ in host defense. To explain the benefits of enteral nutrition, a breakdown of the gastrointestinal barrier during parenteral nutrition is hypothesized as a source of inflammatory molecules and bacteria gaining entry into the body that may influence the development of multiple organ failure, sepsis, and the sepsis syndrome. Investigators have focused on several aspects of gut-barrier function, including integrity of the mucosa, [7,18] changes in intestinal permeability,  changes in intestinal bacterial flora, and the translocation of bacteria to MLNs, and the interrelationships of these factors. [3-6,12] Most studies of GALT function have focused on concentrations of IgA recovered from the intestinal or biliary tract and, with few exceptions, have ignored the GALT cell population changes. [9,19,20] The current study demonstrates that TPN solution administered intravenously or enterally results in significant atrophy of the GALT throughout the LP, PPs, and IE spaces with a decrease in the CD4/CD8 ratio in the LP. Our results show a correlation between this generalized atrophy and decreased levels of intestinal intraluminal IgA. These changes can be minimized or eliminated when a complex enteral formula is administered enterally, but the atrophy and decreased GALT function are not reduced when an elemental formula is administered via the gastrointestinal tract.
The GALT can be divided into two separate but interconnected components: the affector site (PPs and MLNs) initiates contact between native B cells and T cells, and intraluminal antigen that has been taken up by M cells and processed by antigen-presenting cells. The effector limb consists of the LP T lymphocytes and B lymphocytes, and IE lymphocytes located between the mucosal epithelial cells.  Antigen uptake into the GALT results in stimulation of B cells and T cells that leave the affector site via efferent lymphatics and home to the effector sites of the gastrointestinal tract, as well as the upper respiratory tract, reproductive tract, and exocrine glandular tissues throughout the mucosal-associated lymphoid tissue system.  The LP within the gut is the major mucosal effector site and is enriched with IgA-positive plasma cells that are covered by epithelial cells producing the polymeric immunoglobulin receptor to transport polymeric IgA to be released at the epithelial apical surfaces into the intestinal lumen. [21-23] Approximately 80% of the plasma cells in the small intestine produce antibodies of the IgA isotype.  With continued antigenic stimulation, IgA production produces an immunologic barrier to the specific antigens on bacteria and bacteria products, food antigens, and so on.
IgA plays an important role in intestinal immune defenses, because it prevents the adherence of bacteria to mucosal epithelial cells, an event considered to be crucial in initiating mucosal invasion and infection.  Experimentally, IgA deficiency results in adherence of intraluminal bacteria to epithelial cells, with subsequent translocation. Previous investigators have characterized relationships between the route and type of nutrition, biliary or intestinal IgA levels, bacterial overgrowth, and bacterial translocation to mesenteric lymph nodes. [3-6,9,20] IV-TPN depresses biliary IgA, increases gramnegative and aerobic bacterial concentrations in the cecum, and increases bacterial translocation. Bacterial translocation is somewhat reduced when TPN solution is given enterally, showing an effect of route of nutrient delivery when compared with intravenous administration of the same solution, but reduced even more when the elemental diet is supplemented with fiber, illustrating the impact of type of nutrient administration on host defense.  Amounts of IgA necessary for adequate host defense may vary depending on conditions and the observation by Spaeth et al.  that increases in gramnegative enteric bacterial and total aerobes in intravenous-fed animals resulted in bacterial translocation with no change in IgA. This suggests an IV-TPN-induced depression, because intestinal IgA normally increases in response to increases in bacterial concentrations.
The consistent relationship between decreased T-cell and B-cell yield and decreased IgA recovery in our study suggests an interdependency and potentially immunosuppressive effect of intravenous or elemental enteral feeding. The cell phenotype distribution in control animals is consistent with other published work, [25-29] and our data implicate the decrease in GALT cellularity and changes in cell phenotype as important factors in decreased intestinal, but not biliary, IgA levels in mice. Cellularity changes were noted in almost all GALT compartments. B cells and T cells were significantly reduced in PPs, which serve as the precursor source of T cells and B cells destined to repopulate the LP for control and production of IgA synthesis. After 5 days of feeding, there was a 55 to 66% reduction in LP B cells with enteral or parenteral feeding of TPN, compared with chow and a 40% reduction in the number of T cells. The reduced B-cell and T-cell numbers paralleled the 45 to 49% decrease in total intestinal IgA and confirm previous observations of a decrease in IgA-positive cells and generalized atrophy within ileal LP associated with intravenous feeding, using immunohistochemical techniques. [9,19] The difference between intestinal and gallbladder values in our study is not surprising, because the source of intestinal IgA in mice (and humans) is primarily across the epithelium through synthesis locally within the LP, [23,29,30] potentially explaining normal gallbladder (and presumably biliary) IgA levels that normally drop with TPN in rats who obtain most intestinal IgA from biliary origin.
The decreases in intestinal IgA also correlated with a shift toward CD8sup + cells. Antigenic challenge elicits B-cell development and IgA-dominated response through an elaborate and highly regulated process that is involved with regulation of the mucosal immune system. Affector and effector sites contain distinct T-cell subsets that possess unique biologic characteristics for the induction and regulation of IgA immune responses. In our study, approximately 40% of the LP lymphocytes in control animals were CD3sup + cells, with the majority being CD4sup + /CD8sup + and the CD4:CD8 ratio approximated 1.7/1. These values were preserved in animals fed complex enteral diets, but the CD4:CD8 ratio dropped in the IG-TPN-fed and IV-TPN-fed animals, primarily because of drops in the CD4sup + cells. Potentially, a shift in balance of cytokine production by these cells could occur. Normally, CD4sup + cells can be divided into two classes, Th1 and Th2, on the basis of cytokine secreted.  Characteristically, Th1 cells secrete interleukin (IL)-2, interferon-gamma, and tumor necrosis factor-beta (lymphotoxin), whereas Th2 cells secrete IL-4, IL-5, IL-6, and IL-10. Although functionally both of these CD4sup + cell types stimulate B cells to proliferate, undergo colony expansion, and secrete immunoglobulin when exerting positive effects on IgA secretion,  Th2 cells are generally considered more effective in stimulating IgA production. Whether there is a shift from Th2 to Th1 cells is unknown, but there clearly is a decrease in CD4sup + cells that could exert negative effects on an already decreasing mass of B cells in the LP.
Causes for the reduction in the total lymphocyte pool in the LP were not investigated in the current experiments. However, we do not feel that this reduction is caused by TPN-induced generalized intestinal atrophy, because parenteral feeding does not produce villus atrophy in the mouse model.  Reduction in the number of recirculating lymphocytes from PPs through MLNs to the LP or changes in homing are possibilities, because decreases in lymphocyte content in PPs correlated well with decreases in the LP. Although the origin and fate of IE lymphocytes are not fully understood, there is the possibility that these lymphocytes (which are primarily CD8sup + ) could regulate IgA-producing cells in the LP, exerting inhibitory effects as suggested by some investigators.  Most of the changes could be avoided by the use of a complex, nonfiber-containing diet. Perhaps this diet stimulated increased levels of hormones or neuropeptides that are postulated to be important in GALT function. [35-37] Recently, neurotensin receptors were found on peripheral lymphocytes, suggesting the role of this neuropeptide on the GALT.  Others have shown effects of gastrin, cholecystokinin, and bombesin on preservation of intestinal mucosa, bacterial translocation, or intestinal IgA levels. [36,39]
The current study provides insight into the importance of route of nutrient administration on immunologic defenses in critically ill and critically injured patients. Preservation of the GALT with enteral feeding may result in immunologic integrity of the mucosal-associated lymphatic tissue and increased resistance to infection. The ability to manipulate mucosal defenses via this model may allow the development of therapies that can preserve immunologic function in patients unable to be fed via the gastrointestinal tract.
We thank Doris Parsons for her invaluable help in preparing this manuscript.
(Dr. Ed Deitch's remarks were read by Dr. Livingston.)
Dr. David H. Livingston (Newark, New Jersey): Dr. Deitch called me last night, indicating he could not come to the meeting and sends his regrets for not being able to discuss this very nice study. He sent me his comments which I will relay.
This is a very nice and technically very sound study with very good methodology. It shows that total parenteral nutrition solution administered intravenously or enterally clearly alters intestinal immune system or the gut-associated lymphatic tissue (GALT), as it was termed, in a number of ways. Because the study has really no technical problems, we wish to focus the discussion on the interpretation of the work rather than the methodology.
The basic implication of this work is that impaired function of GALT, in one way or another, is important in nutritionally induced gut barrier failure and subsequently infection. How accurate is this assumption?
Individuals from his own laboratory and others have demonstrated that parenteral alimentation, as well enteral alimentation, have clearly promoted bacterial translocation experimentally. We also know that these diets cause a number of morphologic and functional changes in the intestine itself, as well in the GALT. And, thus, you have clearly shown that changes in the GALT do occur.
Do you have any direct evidence to show that these changes in intestinal immunity, rather than some other unmeasured nutritionally induced changes in function, are causally related to gut origin sepsis?
Data from Dr. Deitch's laboratory would suggest that diet induced bacteria translocation is caused more by the effects of diet than the immune system. In fact, they have observed diet-induced decreases in intestinal secretory IGA. However, therapeutic maneuvers that successfully restore intestinal antibacterial barrier function have not been associated with improvements in immunoglobulin A (IgA) levels to normal. Could you comment on this?
Next, the clearest and perhaps most important conclusion to be drawn from this data, as well as a preponderance of other data studied in this area of nutrition, is that nothing is better for rats than rat chow. Therefore, by implication, nothing is better for humans than human chow. Why is this? Is it because now chow contains some magical compounds that diets we test and give our patients are deficient in some essential ingredient?
Our studies have shown that the magical compound in rat chow that may be missing in our human diets is not from animal fiber. In fact, recent studies from our laboratory indicate that fiber is protective against, translocation through the mediation of induction of trophic hormones and that intestinal barrier function can be modulated hormonally.
I would be interested in knowing your thoughts about why rat chow is so good.
Lastly, we were not able to determine from the paper exactly how you processed the IgA levels. Secretory IgA can bind to bacteria. How do you know that the low levels of IgA observed are not caused by bacterial binding?
In conclusion, we wish to congratulate the authors on a very excellent, elegant study that provides us with new and important information.
Dr. Fredrick A. Moore (Denver, Colorado): The gastro-intestinal tract is now recognized to be an important immunologic organ both locally and systemically. You have shown that feeding a total parenteral nutrition (TPN) solution intravenously or via the gut impairs the local gut immunity. Do you plan to measure systemic immune responses? Next, many basic laboratory models comparing enteral to parenteral nutrition use TPN solutions as the study diet. They then equate TPN solutions to the enteral elemental diets use clinically. This is not a fair comparison. For example, TPN solutions have a very high osmolality, which is most likely quite toxic to the gut mucosa when fed enterally. Could you comment on this issue?
Dr. Kenneth A. Kudsk (closing): I would like to thank both Dr. Livingston and Dr. Moore for their comments.
Is there any direct evidence of impaired immunity related to the gut origin sepsis? Well, I think the gut origin sepsis theory is still extremely controversial. We did not look at bacterial translocation, or any translocation of bacterial products across the mucosa in this particular animal model, because we were checking the mesenteric lymph nodes for cell population changes that might occur.
I think that from data that we have, as well as data that has been obtained from burn patients by Dr. Deitch and Drs. Ziegler and Wilmore, there certainly is an increase in permeability in the intestine that occurs following burns. In a study from our institution, within 72 hours of injury, there is a great increase in intestinal permeability to lactulose and mannitol. It is very conceivable that there are products other than the bacteria that could cross the mucosa having downstream effects.
In fact, we have seen a relationship between interleukin-6 secretion from the gastrointestinal tract, a proinflammatory molecular that affects acute phase protein production by the liver, and an increase in gut permeability.
I did not quite understand the second question about immunoglobulin A (IgA), so I will have to pass on that.
Is rat chow best and what are the factors that make it best? Well, I think there is a number of things. Part of it is composition; things like polyamines and other microsubstances are found in general nutrition, but are not going to be found in these chemically defined diets.
But there is another important effect and that is hormonal stimulation. When these animals eat chow, they have intermittent boluses of hormonal stimulation to the gastrointestinal tract. We have just completed a study using a neuropeptide in which the animals are fed the elemental diet both with and without this hormonal neuropeptide. We have partially, if not completely, reversed the atrophy that we noted in this particular study.
Could our decreased IgA levels be secondary to bacteria? We did not measure bacteria levels in the gastrointestinal tract in the initial experiment. We did start doing that in the second experiment. We did not find bacterial overgrowth. So, I doubt that lower IgA levels are secondary to bacterial overgrowth and increased IgA binding.
And, finally, Dr. Moore asked about the systemic versus local immunity. You have to realize that 50% of the immune system of the body is located along the gastrointestinal tract. Eighty percent of the humoral antibody that the body makes goes across the gastrointestinal tract to act as a barrier. This intestinal immunity has largely been ignored in previous studies. Investigators looked at IgA and at bacterial translocation. But everybody has pretty much ignored this area.
You have to consider the gut, the epithelial immune system as a systemic, total mucosal immune system as a systemic, total mucosal immune system. Within the mouse, it has been very well worked out. When you give a challenge to the gastrointestinal tract, shortly thereafter, within 10 to 14 days, specific IgA against that antigen is found in the nasopharynx, within the lungs.
We need to think of it as a mucosal immune system with cells from the gut repopulating other areas of the body, providing protection to various antigens that the gut has seen.
In terms of using total parenteral nutrition as a model of an elemental diet, it is sort of a catch-22. If you use another kind of elemental diet and compare it with total parenteral nutrition, there is the criticism, ``Oh, you are not using identical solutions.'' So, the best way, I guess, to get to the bottom line is to use identical solutions enterally and parenterally as standards. Thank you.
1. Salmi M, Jalkanen S: Regulation of lymphocyte traffic to mucosa-associated lymphatic tissues. Gastroenterol Clin North Am 20:495, 1991
2. Ottaway CA: Neuroimmunomodulation in the intestinal mucosa. Gastroenterol Clin North Am 20:511, 1991
3. Spaeth G, Berg RD, Specian RD, et al: Food without fiber promotes bacterial translocation from the gut. Surgery 108: 240, 1990
4. Deitch EA, Xu D, Qi L, et al: Elemental diet-induced immune suppression is caused by both bacterial and dietary factors. JPEN 17:332, 1993
5. Alverdy JC, Aoys E, Moss GS: Total parenteral nutrition promotes bacterial translocation from the gut. Surgery 104:185, 1988
6. Mainous M, Xu D, Lu Q, et al: Oral-TPN-induced bacterial translocation and impaired immune defenses are reversed with refeeding. Surgery 110:277, 1991
7. Li J, Langkamp-Henken B, Suzuki K, et al: Glutamine prevents parenteral nutrition-induced increases in intestinal permeability. JPEN 18:303, 1994
8. Purandare S, Offenbartl K, Westerom B, Bengmark S: Increased gut permeability to fluorescein isothiocyanatedextran after total parenteral nutrition in the rat. J Gastroenterol 24:678, 1989
9. Alverdy JC, Aoys E, Weiss-Carrington P, et al: The effect of glutamine-enriched TPN on gut immune cellularity. J Surg Res 52:34, 1992
10. Mahaffey SM, Copeland EM, Talbert JL, et al: Parenteral nutrition in the Swiss mouse. Fed Proc 43:848, 1984
11. National Academy of Science: Nutrient Requirements of Laboratory Animals. Washington, DC, National Academy of Science, National Research Council Publication No. 10, 1978
12. Deitch EA, Xu D, Qi L: Different lymphocyte compartments respond differently to mitogenic stimulation after thermal injury. Ann Surg 211:72, 1990
13. Kudsk KA, Croce MA, Fabian TC, et al: Enteral vs parenteral feeding: Effects on septic morbidity following blunt and penetrating trauma. Ann Surg 215:503, 1992
14. Moore FA, Moore EE, Jones TN: Benefits of immediate jejunostomy feeding after major abdominal trauma--A prospective randomized study. J Trauma 26:874, 1986
15. Moore FA, Moore EE, Jones TN, et al: TEN vs. TPN following major abdominal trauma--Reduces septic morbidity. J Trauma 29:916, 1989
16. Kudsk KA, Stone JM, Carpenter G, et al: Enteral and parenteral feeding influences mortality after hemoglobin E. coli peritonitis in normal rats. J Trauma 23:605, 1983
17. Kudsk KA, Carpenter G, Petersen SR, et al: Effective enteral and parenteral feeding in malnourished rats with hemoglobin-E. coli adjuvant peritonitis. J Surg Res 31:105, 1981
18. Johnson LR, Copeland EM, Dudrick SJ, et al: Structural and hormonal alterations in the gastrointestinal tract of parenteral fed rats. Gastroenterology 68:1177, 1975
19. Tanaka S, Miura S, Tahiro H, et al: Morphological alteration of gut-associated lymphoid tissue after long-term total parenteral nutrition in rats. Cell Tissue Res 266:29, 1991
20. Spaeth G, Gottwald T, Haas W, et al: Glutamine peptide does not improve gut barrier function and mucosal immunity in total parenteral nutrition. JPEN 17:317, 1993
21. Kiyono H, McGhee JR: T helper cells for mucosal immune responses. In Ogra PL, Lamm ME, McGhee JR, et al. (eds): Handbook of Mucosal Immunology. New York, Academic Press, 1994, pp 263-273
22. Brandtzaeg P: The role of J chain and secretory component in receptor-mediated glandular and hepatic transport if immunoglobulins in man. Scand J Immunol 22:111, 1985
23. Ofek I, Beachey EH: Bacterial adherence. Adv Intern Med 25:503, 1980
24. Mestecky J, McGhee JR: Immunoglobulin A (IgA): Molecular and cellular interactions involved in IgA biosynthesis and immune response. Adv Immunol 40:153, 1987
25. Kiyono H, McGhee JR: T helper cells for mucosal immune responses. In Ogra PL, Mestecky J, Lamm ME, et al. (eds): Handbook of Mucosal Immunity. New York, Academic Press, 1994, pp 263-274
26. Kiyono H, McGhee JR, Mosteller LM, et al: Murine Peyer's patch T-cell clones: Characterization of antigen specific helper T cells for IgA response. J Exp Med 79:596, 1982
27. Maloy KJ, Mci A, Zamoyuska, R, et al: Phenotypic heterogeneity of intraepithelial T lymphocytes from mouse small intestine. Immunology 72:555, 1991
28. Mosley RL, Klein JR: A rapid method for isolating murine intestine intraepithelial lymphocytes with high yield and purity. J Immunol Methods 156:19, 1992
29. Harriman GR, Hornqvist E, Lycke NY: Antigen-specific and polyclonal CD4sup + lamina propria T-cell lines: Phenotypic and functional characterization. Immunology 75:66, 1992
30. Delacroix DL, Malburny GN, Vaerman JP: Hepatobiliary transport of plasma IgA in the mouse: Contribution to the clearance of intravascular IgA. Eur J Immunol 15:893, 1985
31. Mossmann TR, Coffman RL: Th1 and Th2 cells: Different patterns of lymphokine secretion lead to different functional properties. Annu Rev Immunol 7:145, 1989
32. Coffman RL, Seymour BHP, Lebman DA, et al: The role of helper T cell products in mouse B cell differentiation and isotope regulation. Immunol Rev 102:5, 1988
33. Sitren HS, Bryant M, Ellis LM: Species differences in TPN-induced intestinal villus atrophy. JPEN 16(suppl): 30S, 1992
34. Taguchi T, Aicher WK, Fujihashi K: Novel function for intestinal intraepithelial lymphocytes. J Immunol 147:3736, 1991
35. Haskel Y, Xu D, Lu Q, et al: Bombesin protects against bacterial translocation induced by three commercially available liquid enteral diets: A prospective, randomized, multigroup trial. Crit Care Med 22:108, 1994
36. Haskel Y, Xu D, Lu Q: Elemental diet-induced bacterial translocation can be hormonally modulated. Ann Surg 217:634, 1993
37. Haskel Y, Xu D, Lu Q, et al: The modulatory role of gut hormones in elemental diet and intravenous total parenteral nutrition-induced bacterial translocation in rats. JPEN 18:159, 1994
38. Evers BM, Izukura M, Townsend CM, et al: Neurotensin prevents intestinal mucosal hypoplasia in rats fed an elemental diet. Dig Dis Sci 37:426, 1992
© Williams & Wilkins 1995. All Rights Reserved.
39. Jin G-F, Guo Y-S, Houston CW. Bombesin: An activator of specific aeromonas antibody secretion in rat intestine. Dig Dis Sci 34:1708, 1989