Significant clinical and experimental evidence shows that severely injured patients fed enterally incur less infectious morbidity compared to patients fed parenterally or starved. In studies of severely injured trauma patients, individuals randomized to enteral feeding sustained significantly fewer pneumonias and intraabdominal abscesses than an injured group fed parenterally [15
]. Experimentally, our work and that of others show that route and type of nutrition influence established mucosal immunologic defenses [13
], intraperitoneal cytokine and immunologic responses [18
], and systemic inflammatory responses to injury [6
]. This work focuses on PN induced alterations in mucosal immunity.
Mucosal immune responses are initiated when luminal antigen is taken up by M cells on the surface of Peyer's patches and processed within this structure leading to activation of T & B cells. Naïve T and B lymphocytes are attracted to the Peyer's patches by a specific adhesion molecule, mucosal addressin cellular adhesion molecule-1, and are sensitized to the processed antigen. The lymphocytes migrate through mesenteric lymph nodes, the thoracic duct and back into the blood stream and home to the lamina propria in the respiratory tract and intestine (as well as all other mucosal sites) where the mature lymphocytes produce dimeric IgA [14
] The IgA binds to pIgR located on the basolateral surface of epithelial cells. pIgR has 7 domains: 5 extracellular, 1 transmembrane, and 1 cytoplasmic. This pIgR-IgA complex is endocytosed by epithelial cells, and transported to the apical cell surface. The complex is released into the lumen as secretory IgA by cleavage of the extracellular part of pIgR, also called secretory component [12
]. Secretory IgA prevents invasive infection by blocking pathogen attachment to mucosal surfaces.
Our work consistently shows that chow feeding maintains normal mucosal defenses but alterations in route and type of nutrition significantly affect MALT and GALT with direct effects on the main antigen-specific mucosal immune defense, IgA. Parenteral feeding (with no enteral stimulation) significantly reduces intestinal and respiratory defenses by decreasing levels of molecules which are responsible for directing T & B cells into and through the mucosal immune network [9
]. Such changes contribute to impaired mucosal immune function and integrity through several means including (among others) IgA production and IgA transport. We documented that PN decreases the number of T & B cells in the intestine [17
] and lung (unpublished data), reduces the number antibody-forming cells in the upper respiratory tract (reduced production) in response to an acute infection [11
], reduces levels of Th2-type IgA stimulating cytokines (reduced production), such as IL-4 and IL-10 [31
], and reduces levels of pIgR protein within the intestinal mucosa (reduced transport) [26
]. The overall effect of these alterations reduce available IgA production (reduced T & B cells and Th2-type cytokines) and transport (lowered intestinal pIgR levels) to impair or eliminate established anti-viral and anti-bacterial defenses in our murine model [13
]. In one previous experiment, however, we documented that respiratory protection against the A/PR8 (H1N1) influenza virus was preserved in PN-fed mice when influenza-specific monoclonal polymeric IgA was provided exogenously [25
]. This suggested that airway transport was still adequate to provide protection if adequate IgA was provided and led to our hypothesis that PN induces unique tissue-specific effects on transport and production.
The present work confirms this hypothesis by quantifying levels of IgA and pIgR in tissue homogenates, and luminal IgA in the intestine, lung and upper respiratory nasal passages. Lumen lavage specimens quantify the levels of IgA secreted onto mucosal surfaces by the epithelial cells while tissue homogenate IgA reflects IgA available for transport by pIgR, and pIgR protein levels measure an important component of the transport process. IgA levels in lavage specimens from the nasal passages and intestine dropped significantly with PN compared to chow fed mice replicating prior observations in regard to this particular outcome [25
] and assuring that the model functioned and reliably reproduced conditions to interpret the new findings. Uniquely, this work demonstrates organ specific
effects of enteral stimulation - or lack of enteral stimulation - on tissue (not luminal) IgA levels and the critical IgA transport protein, pIgR. Tissue homogenate IgA levels drop significantly in lung and nasal passages of PN-fed mice, but levels are maintained in the small intestine. PN-fed mice preserve pIgR expression in the nasal passage and seem to augment pIgR levels in the lung consistent with our observations that adequate residual transport is present in the upper respiratory tract when IgA is provided exogenously [25
]. Our data implies that PN has less effect on respiratory transport (although we limited our measurement to only one component of transport) than on production. This is consistent with our observations that acute infection leads to lower numbers of IgA producing cells in the nasal passages of PN-fed mice [11
] and the low levels of tissue homogenate IgA found in our specimens. Conversely, intestinal pIgR levels drop with PN, while SI tissue homogenate IgA stays unchanged. The result of decreased pIgR expression in small intestine after PN feeding remains consistent with our hypothesis because small intestinal levels of IL-4 (a Th-2 cytokine that stimulates pIgR production) decreases in PN-fed mice [31
]. This almost certainly implicates impaired transport as a main factor in the decrease of SI mucosal surface IgA levels with PN.
There are several limitations to this study. First, the IgA compartments (luminal by lavage, and tissue by homogenization) cannot be completely separated. We investigated the effect of the initial lavage on levels of IgA remaining in the lumen by measuring IgA in sequential washings of chow fed mice to address this problem. The initial lavage captured the majority of the IgA collected by this method in the intestine and lung with the amount of remaining luminal IgA small compared to levels measured in the tissues (< 5% in intestine and < 2% in lung). This residual surface IgA is unlikely to significantly falsely elevate such measurements particularly in PN fed mice that have even lower luminal IgA levels than the chow fed mice. Secondly, we cannot sub-localize “tissue” IgA to the lamina propria (production) or to the epithelial cell layer (in transit) with current techniques. Without such an assay, the intestinal tissue homogenate IgA results should be interpreted with caution but we are reluctant to believe that IgA production would trend
towards increase in PN animals given our previous data showing dramatic, 50% reductions in T & B cell numbers in the gut lamina propria [17
] and decreases in IL-4 and IL-10, two important IgA stimulating cytokines, with PN [31
]. Such data would suggest that IgA production is profoundly affected in the small intestine with PN. We believe this trend towards increased tissue IgA may be partially an artifact since our results are expressed as IgA mass/mg protein. We recently noted in a pilot study that total small intestinal protein dropped by almost 40% in PN vs. chow-fed mice (unpublished data). Lavage IgA results are not subject to such influences since they are measured in mass/volume of lavage effluent and the lavage volume is constant for all animals. A specific ELISA for detecting murine secretory component could potentially aid in solving this problem of IgA “sub-localization” but currently such an assay is unavailable since the anti-SC antibody used for Western blotting does not work in an ELISA. However, we acknowledge the potential value of immunohistochemical studies to complement the current work and are developing such assays after quantifying this process using whole tissue samples. Traditional cross sectional immunohistochemical analysis may be problematic to study events in the lamina propria [32
] and we are approaching such analysis with caution.
In the context of previously published work, these results continue to develop mechanistic explanations for the primary end effects of PN feeding on respiratory immunity; namely, decreased mucosal IgA levels and an inability to generate new immunity or utilize previously acquired responses in the respiratory tract. This work also provides evidence that alterations in type and route of nutrition alter the mucosal immune components of respiratory and intestinal organ systems in different ways. Clinically this may prove an important distinction and future efforts will be designed to discern the basic mucosal immune alterations of IgA production and transport that are organ/tissue specific
. While mucosal immune lymphocytes are thought to be educated in the gut before being distributed throughout the body, local signals at their final destination likely effect their terminal function. We previously showed that chemokine mRNA and protein levels are altered in a tissue specific manner with PN [33
]. Cytokine profiles have also been shown to be tissue specific [34
] which likely influences immune cell function within those tissues.
In summary, reduction in pIgR expression appears primarily responsible for reduced secretory IgA levels in the small intestine of PN-fed mice. On the other hand, PN appears to reduce IgA production rather than transport (at least via pIgR) in the respiratory tract.