Airway-specific recruitment of T cells is reduced in a CD26-deficient F344 rat substrain


S. von Hörsten, Experimental Therapy, Franz-Penzoldt-Center, Friedrich-Alexander-University Erlangen-Nürnberg, Palmsanlage 5, 91054 Erlangen, Germany.


Asthma is a chronic inflammatory disease affecting the airways. Increased levels of T cells are found in the lungs after the induction of an allergic-like inflammation in rats, and flow cytometry studies have shown that these levels are reduced in CD26-deficient rats. However, the precise anatomical sites where these newly recruited T cells appear primarily are unknown. Therefore, we quantified the distribution of T cells in lung parenchyma as well as in large, medium and small airways using immunohistochemical stainings combined with morphometric analyses. The number of T cells increased after the induction of an allergic-like inflammation. However, the differences between CD26-deficient and wild-type rats were not attributable to different cell numbers in the lung parenchyma, but the medium- and large-sized bronchi revealed significantly fewer T cells in CD26-deficient rats. These sites of T cell recruitment were screened further using immunohistochemistry and quantitative real-time polymerase chain reaction with regard to two hypotheses: (i) involvement of the nervous system or (ii) expression of chemokines with properties of a T cell attractor. No topographical association was found between nerves and T cells, but a differential transcription of chemokines was revealed in bronchi and parenchyma. Thus, the site-specific recruitment of T cells appears to be a process mediated by chemokines rather than nerve–T cell interactions. In conclusion, this is the first report showing a differential site-specific recruitment of T cells to the bronchi in a CD26-deficient rat substrain during an asthma-like inflammation.


Asthma is a chronic inflammatory disease of the airways, which is characterized mainly by bronchial hyperresponsiveness, airway obstruction and airway inflammation with an influx of different cell types. The inflammation causes wheezing, coughing, chest tightness and breathlessness and is accompanied by an elevated production of cytokines and chemokines, as well as oedema and mucus hyper-secretion (for review see [1]). An immunoglobulin E (IgE)-mediated response very similar to the response in human bronchial asthma can be provoked in different animal models using inhaled allergen challenge. Mice, rats, guinea pigs and other species, as well as different substrains of species, have been established as animal models. Each model has its own advantages, but no model is identical to the conditions found in human disease [2]. Rats as well as other rodents exhibit a monopodial bronchial branching pattern, which is fundamentally different from that of humans, but the more human-like structure of the submucosal and bronchial smooth muscle layers and the existence of bronchial arteries in addition to capillaries and veins in the lamina propria of the bronchi make the rat a good model species for bronchial asthma [3].

In a Fischer 344 (F344) rat model for asthma, we have already documented a dose-dependent recruitment of T cells into the lung [4]. T lymphocytes are thought to play a pivotal role in the pathogenesis of asthma and many disease parameters can be modulated by T cell recruitment and activation, cell adhesion and chemokine metabolism. All these disease-modulating processes are potentially influenced further by CD26 expression and its associated enzymatic activity [5]. CD26 is a multi-functional glycoprotein, which is also called dipeptidyl peptidase 4 (DP4) (DPP4, DPPIV, EC It exists in a membrane-bound form as well as in soluble form, and is capable of regulating various physiological processes by N-terminal cleavage of dipeptides from peptides with L-proline/-alanine at the penultimate position [6] including, but not limited to, chemokines such as stromal cell-derived factor-1 (SDF-1, CXCL12). The lung has been described as the organ with the second highest dipeptidyl peptidase activity of CD26 [7], while the bronchi do not express CD26 [8].

However, at the sites of the bronchi a structural homologue of CD26, DP10, is expressed [8]. The gene of DP10 represents a susceptibility locus for asthma in humans [9], and its protein is associated with Kv4-mediated A-type potassium channels [10] that play a pivotal role controlling cell excitability, especially in the nervous system. In addition to DP10, the low-affinity pan-neurotrophin receptor p75NTR is expressed in airways [11]. The p75NTR is also expressed on many different cell types in the nervous system and can mediate a variety of different cellular functions, including neuronal survival, cell death and neurite growth. Recent data also indicate an involvement of neurotrophins in asthma via their binding to p75NTR[12].

An active-site mutation of CD26 in a Fischer rat substrain has caused the retention and degradation of CD26 in the endoplasmic reticulum of these rats [13]. The induction of an allergic-like inflammation in this substrain has shown differences in the recruitment of T cells compared to a CD26-expressing wild-type substrain [14]. These differences were based on flow cytometry analyses of the bronchoalveolar lavage (BAL) of F344 rat substrains, but morphometric analyses of the distribution of T cells in different compartments of the lungs are still lacking.

The present study investigates the distribution of T cells in lungs of CD26-deficient and wild-type F344 rat substrains under naive conditions and after the induction of an allergic-like inflammation, with regard to the two hypotheses of either nervous system involvement or a CD26-dependent and chemokine-mediated compartment-specific differential recruitment.

Materials and methods


Male wild-type F344/Ztm rats (CD26pos) and male CD26-deficient mutant rats [F344/Crl(Wiga)SvH-Dpp4m] lacking DP4 activity as well as DP4/CD26 expression (CD26neg) were used [15]. All animals were housed at the Central Animal Facility (Ztm) of the Hannover Medical School, maintained in a separate minimal barrier-sustained facility, and monitored microbiologically according to Federation of European Laboratory Animal Science Associations recommendations [16]. All research and animal care procedures had been approved by the review board of the Landesamt fuer Verbraucherschutz und Lebensmittelsicherheit (LAVES; Oldenburg, Germany), and were performed according to international guidelines on the use of laboratory animals.

Sensitization and allergen challenge

Two repeated experiments were performed, first with five animals per group, later with four animals per group. At the age of 6 months, CD26-positive and CD26-negative rats in the asthma group were sensitized 14 and 7 days before challenge, as described previously [4]. In brief, in each case sensitization was performed with 1 mg of ovalbumin (OVA) (Sigma, Deisenhofen, Germany) and 200 mg of Al(OH)3 (Sigma) in 1 ml 0·9% (sterile, pyrogen-free) NaCl injected subcutaneously into a hind-limb. As a second adjuvant, concentrated preparations of 6 × 109 heat-killed Bordetella pertussis bacilli (kindly provided by Chiron Behring, Marburg, Germany) in 0·4 ml 0·9% NaCl were given intraperitoneally at the same time. Animals were challenged with 7·5% of aerosolized OVA using a Pari LC Star nebulizer (Pari, Starnberg, Germany). Three CD26-positive and three CD26-deficient rats were neither sensitized nor challenged, serving as a control group.

Dissection of animals

The animals were dissected under isoflurane anaesthesia 22 ± 0·5 h after challenge, as described previously [4]. Briefly, the animals were killed by aortic exsanguination and blood samples were collected. For BAL isolation, a cannula was inserted into the trachea in situ and the lungs were lavaged four times with portions of 5 ml 0·9% NaCl solution. The recovery of fluid from all animals was greater than 90%. For further analysis, the lungs were excised. Whole left lungs were instilled with 3 ml of Tissue-Tek® O.C.T. compound (Miles Inc., Elkhart, IN, USA) mixed 1 : 4 with phosphate-buffered saline and placed on aluminium foil on dry ice. For PCR analyses, the large bronchi and the parenchyma of the right lungs were excised and frozen in liquid nitrogen.

Differential cell counts

Total cell numbers of the BAL were determined in a Neubauer counting chamber (Hecht, Sondheim, Germany) and differentials were obtained on cytospots using Quick-Diff, as described previously [14].

OVA-specific IgE enzyme-linked immunosorbent assay

OVA-specific IgE levels in plasma obtained from peripheral blood were determined by enzyme-linked immunosorbent assay, as described previously [14].


Analysis of the compartmentalization of T cells, T cell subpopulations and components of the nervous system in the lung was performed using monoclonal (mAbs) and polyclonal (pAbs) antibodies together with different staining methods.

Two consecutive alkaline phosphatase anti-alkaline phosphatase stainings [17] were performed on 10 µm acetone-fixed cryostat sections of the whole left lungs with Fast Blue (Sigma) or Fast Red (Sigma) as the detection system for labelled cells, as described previously [8] with 30 min incubation for T cells (mouse mAb R73; AbD Serotec, Duesseldorf, Germany), CD4 (mouse mAb W3/25; AbD Serotec), CD25 (mouse mAb OX39; AbD Serotec) and overnight incubation for CD26 (mouse mAb 5E8; Hycult Biotechnology b.v., Uden, the Netherlands). Sections were counterstained with hemalaun (1 : 5 in phosphate-buffered saline; Merck, Darmstadt, Germany) for 20 s and covered with Mowiol (Hoechst AG, Frankfurt/Main, Germany).

In addition, immunofluorescence histochemistry was perfomed with a primary pAb raised in rabbits for p75NTR (Millipore, Schwalbach, Germany) together with a primary mouse mAb R73 (AbD Serotec). Rabbit or mouse IgGs in appropriate dilutions were used instead of the primary antibodies as isotype controls. Cy3-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch, Suffolk, UK) and Alexa Fluor 488-conjugated donkey anti-mouse IgG (Invitrogen, Karlsruhe, Germany) were used as secondary antibodies. Counterstaining with Hoechst dye (Invitrogen) was performed in all specimens and the sections were covered with Mowiol (Hoechst AG).

Quantitative histology of lung tissues

For compartmentalization of T cells, the optical dissector method [18] was chosen during the first series of experiments to determine the absolute cell number in the lungs. Therefore, three acetone-fixed cryostat sections (thickness 40 µm, interval 800 µm) of each rat were evaluated with regard to CD26-expressing and CD26-non-expressing T cells in the different compartments of the lung [parenchyma, bronchi, vessels, alveolar space, perivascular space and bronchus-associated lymphoid tissue (BALT)]. Absolute T cell numbers in the lungs were calculated using a Nikon Eclipse 80i microscope (Nikon GmbH, Duesseldorf, Germany) and the Stereo Investigator software (MicroBrightField Inc., Williston, VT, USA). Test fields were generated across the whole section with a constant interval between the test fields in the x- and y-axes.

In the following series of experiments, the cell density of T cells was determined. Three serial cryostat sections (thickness 10 µm, interval 800 µm) were evaluated of at least four rats per group, as described previously [19]. The cells in the lungs were evaluated using a Nikon Eclipse 80i microscope (Nikon) together with the Stereo Investigator software (MicroBrightField). Test fields were generated across the whole section with a constant interval between the test fields in the x- and y-axes (counting frame: 150 µm × 150 µm; grid size: 1400 µm × 1400 µm) resulting in about 100 test fields per section. T cells were counted and cell density was defined as:


The cells around the bronchi of the lungs were quantified further using a semi-quantitative score (1: < 15 cells; 2: 15–30 cells; 3: > 30 cells). Small bronchi were defined as bronchi with a diameter < 150 µm, medium bronchi were between 150 and 300 µm and large bronchi were defined as having a diameter > 300 µm. At least five bronchi of each size were quantified on all sections, and the mean scores were calculated for each animal.

Representative micrographs were taken with a MicroFire digital microscope camera (Optronics, Goleta, CA, USA).

Quantitative real-time PCR

Tissue samples from shock-frozen bronchi and parenchyma from the lobes of the right lungs were homogenized by means of the homogenizer Precellys with 1·4 mm ceramic beads (5000 rpm, 30 s; Peqlab, Erlangen, Germany). RNA was prepared using the NucleoSpin RNA II kit (Macherey-Nagel, Dueren, Germany) according to the manufacturer's instructions. cDNA preparation and Quantitative real-time PCR (qrtPCR) were performed as described previously [20] with glucose-6-phosphate dehydrogenase (G6pd) as a housekeeping gene (HKG). Briefly, qrtPCR was performed in a Rotor-Gene 3000 (Corbett Research, Hilden, Germany) using the primers displayed in Table 1 as well as the QuantiTect SYBR Green reverse transcription PCR kit (Qiagen, Hilden, Germany). For verification, products' melting curves were generated and single amplicons were confirmed by agarose gel electrophoresis. Relative amounts were determined with the Rotorgene software version 4·6 in comparative quantitation mode.

Table 1.  Primers used for rat G6pd, MCP-1, SDF-1, RANTES, Mig, IP-10, and I-TAC.
GenePrimer sequence (5′-3′)
  1. MCP-1: monocyte chemoattractant protein-1; SDF-1: stromal cell-derived factor-1; G6pd: glucose-6-phosphate dehydrogenase; RANTES: regulated upon activation, normal T cell expressed and secreted; Mig: monokine induced by interferon (IFN)-γ; I-TAC: IFN-inducible T cell α-chemoattractant; IP-10: IFN-γ-inducible protein-10.


Statistical analysis

Differences among groups in T cell density and in the qrtPCR were analysed using two-way anova with treatment (control versus asthma) and genetic background (CD26posversus CD26neg) being the factors, followed by Fisher's tests for protected last significant differences for post-hoc comparisons, if appropriate. Statistically significant effects between the asthma group and the control group are indicated by rhombs (#P < 0·05; ##P < 0·01; ###P < 0·001), and for comparison of CD26neg and CD26pos groups by asterisks (*P < 0·05). All data are displayed as mean ± standard error of the mean (s.e.m.).

The semiquantitative scores of the T cell numbers at large, medium and small airways were analysed with non-parametric statistics for independent samples (Mann–Whitney rank sum test). Statistically significant effects between CD26neg and CD26pos groups are indicated by asterisks (*P < 0·05).


Substrain-specific differences in IgE levels and eosinophil numbers and in allergen-induced T cell peribronchial density, but no significant differences in the lung parenchyma

First, markers of the asthma-like inflammation in the animals were evaluated and resulted in elevated OVA-specific IgE levels after OVA-challenge [wild-type optical density (OD) 2·43 ± 0·47 versus CD26-deficient OD 1·15 ± 0·05] with significantly lower IgE levels in the CD26-deficient rats compared to wild-type rats (*P < 0·05). Eosinophil numbers were also elevated after OVA-challenge (baseline versus challenge; ##P < 0·01) in wild-type animals (baseline 0·0 ± 0·0 versus challenge 1·7 ± 0·6 cells/ml × 106), with a significant lower number of eosinophils in the CD26-deficient rats (baseline: 0·0 ± 0·0 versus challenge: 0·5 ± 0·2 cells/ml × 106) (CD26-positive versus CD26-negative after challenge: *P < 0·05).

The localization of T cells from CD26-positive and CD26-deficient naive F344 rats and rats after the induction of an allergic-like inflammation via an OVA-challenge was determined in six different compartments of the lungs (parenchyma, bronchi, vessels, alveolar space, perivascular space and BALT). The first series of experiments (data not shown) using the optical dissector method showed that the vast majority of T cells were found in the lung parenchyma (up to 90%), and that about 90% of the T cells in wild-type rats were CD26-positive. The absolute T cell numbers increased significantly in the lung parenchyma of both groups after OVA-challenge, but there was no significant difference between wild-type and CD26-deficient rats. A more precise examination of the bronchi in these experiments showed a significantly different T cell number around medium and large bronchi of OVA-challenged wild-type and CD26-deficient animals.

Therefore, the experiments and the immunohistochemical stainings were repeated and the T cell number per mm2 lung was determined in the parenchyma (Fig. 1a). T cells were stained in blue in wild-type (Fig. 1b) and CD26-deficient (Fig. 1c) naive rats. Only a small number of these cells was found in the lung parenchyma. However, the T cell density increased significantly in wild-type (Fig. 1d) and CD26-deficient (Fig. 1e) rats after OVA-challenge. No significant difference was found between wild-type and CD26-deficient rats. CD26 was expressed highly in the lung parenchyma of wild-type rats (Fig. 1b and d), whereas its expression was lacking in CD26-deficient rats (Fig. 1c,e).

Figure 1.

Histological detection of T cells (blue) and CD26 (red) in naive lungs and lungs after ovalbumin (OVA)-challenge. Overview of the T cell density in the parenchyma (a). Representative micrographs of the lung parenchyma of naive wild-type rats (b), naive CD26-deficient rats (c), OVA-challenged wild-type rats (d) and OVA-challenged CD26-deficient rats (e) (###P < 0·001).

The T cell distribution around the bronchi showed differences in the T cell numbers around the large and medium airways of naive and OVA-challenged wild-type and CD26-deficient rats (Fig. 2a–c). In general, the number of T cells in airways increased in both substrains after OVA-challenge. No significant difference in the T cell distribution was observed either in the airways of naive rats from both substrains or in the small airways of both substrains after OVA-challenge. However, a significant difference in the T cell increase around the large and medium airways of wild-type (Fig. 2d) compared to CD26-deficient (Fig. 2e) F344 rats was observed in the OVA-challenged groups.

Figure 2.

Histological detection of T cells (blue) and CD26 (red) around the airways of naive lungs and lungs after ovalbumin (OVA)-challenge. Overview of the T cell numbers around the large airways (a), medium airways (b) and small airways (c). Representative micrographs of the medium airways of wild-type (WT) OVA-challenged rats (d) and CD26-deficient OVA-challenged rats (e); *P < 0·05).

Characterization of cell types in the bronchial environment

The regions of the differential T cell distribution were characterized in more detail. First, CD4+CD25+ cells were stained to determine T cell subsets around the bronchi. Comparison of the four different groups (Fig. 3a,b,d,e) revealed the highest number of these cells around the airways of OVA-challenged wild-type rats. The higher magnifications illustrate CD4-positive, CD25-positive and CD4+CD25+ cells (Fig. 3c,f).

Figure 3.

Histological detection of T cell subpopulations (CD4 in red, CD25 in blue) in the lungs. Representative micrographs of three airways from three wild-type animals from the unchallenged (a) and the ovalbumin (OVA)-challenged group (b), as well as a higher magnification of one airway of an OVA-challenged wild-type animal (c). Representative micrographs of three airways from three CD26-deficient animals from the unchallenged (d) and the OVA-challenged group (e), and a higher magnification of one airway of an OVA-challenged CD26-deficient animal (f) (all bars = 200 µm).

Because nerves and neurotrophins play a critical role during asthma, the expression of the neurotrophin receptor p75NTR was investigated further as a marker for components of the nervous system. The p75NTR-positive cells were found in the BALT of lungs after OVA-challenge in follicular-like areas that were mainly free of T cells (Fig. 4a–c). Such p75NTR-positive structures were also found around the bronchi (Fig. 4d–f), but no differences were detectable between the two substrains.

Figure 4.

Immunofluorescent stainings of T cells (green) and p75NTR (red). T cell receptor (TCR)-positive cells of the bronchus-associated lymphoid tissue (BALT) (a) and p75NTR-positive cells (b) were stained and an overlay (c) was produced. Micrographs of an airway with TCR-positive cells (d), p75NTR-positive structures (e) and an overlay of both (f).

Identification of chemoattractants in the lungs

qrtPCR analyses revealed differential transcript levels of the chemokines monocyte chemoattractant protein-1 (MCP-1) and SDF-1 in the large airways and the lung parenchyma. The transcription of MCP-1 was up-regulated in the large airways (Fig. 5a) and in the lung parenchyma (Fig. 5b) of both OVA-challenged groups. While the transcription of stromal cell derived factor-1 (SDF-1) was up-regulated in the large airways of unchallenged and OVA-challenged wild-type rats compared with CD26-deficient rats (Fig. 5c), it was transcribed equally in the lung parenchyma in all groups (Fig. 5d). The transcriptions of CCL5/regulated upon activation, normal T cell expressed and secreted (RANTES) (Fig. 5e,f) and CXCL11/interferon (IFN)-inducible T cell α-chemoattractant (I-TAC) (Fig. 5k,l) did not show any significant differences, while the transcriptions of CXCL9/monokine induced by IFN-γ (Mig) (Fig. 5g,h) and CXCL10/IFN-γ-inducible protein-10 (IP-10) (Fig. 5i,j) were up-regulated after asthma-induction in both substrains and in both compartments screened. Additionally, the transcription of Mig (Fig. 5h) was significantly higher in the lung parenchyma of the wild-type rats.

Figure 5.

Quantitative real-time polymerase chain reaction data. Relative gene expression of monocyte chemoattractant protein-1 in the large airways (a) and in the lung parenchyma (b), of stromal cell-derived factor-1 in the large airways (c) and in the lung parenchyma (d), of regulated upon activation, normal T cell expressed and secreted in the large airways (e) and in the lung parenchyma (f), of monokine induced by interferon (IFN)-γ in the large airways (g) and in the lung parenchyma (h), of IFN-γ-inducible protein-10 in the large airways (i) and in the lung parenchyma (j), and of IFN-inducible T cell α-chemoattractant in the large airways (k) and in the lung parenchyma (l). (HKG: housekeeping gene glucose-6-phosphate dehydrogenase; #P < 0·05; ##P < 0·01; *P < 0·05).


This study demonstrates blunted IgE levels and eosinophilia in CD26-deficient rats as well as a compartment-specific and differential increase of T cells in an F344 rat model for an asthma-like inflammation.

Immunostaining of T cells in the lungs revealed that their number increases after the induction of this asthma-like inflammation in CD26-positive and CD26-deficient rats. This is consistent with previous flow cytometric analyses on cell suspensions of whole lobes of the right lungs and BAL fluid [4,14]. A reduced increase of T cells in the BAL was reported in CD26-deficient rats after OVA-challenge [14]. However, until now the source of these recruited T cells in the lungs was unknown. Therefore, one aim of the present study was to clarify whether these differentially increased T cells in F344 substrains are distributed equally throughout the whole lungs or whether they accumulate in a certain anatomical compartment of the lungs. After OVA-challenge, staining of whole left lungs revealed an equal distribution of T cells in the lung parenchyma of both substrains, but a significantly different number of T cells between the two substrains around the large and medium airways in the lungs. Surprisingly, this compartment of the lung almost completely lacks CD26 expression as well as DP4 activity [8]. Thus, despite using CD26-deficient and CD26-competent F344 rat substrains, the findings illustrate a difference in an anatomical compartment that largely lacks expression of CD26. This led to the hypothesis that the reason for this phenomenon might be (i) a different environment of the T cells in this area, (ii) the expression or lack of CD26 on the T cells recruited to this site or (iii) other differences, for example in the chemokine pattern at these sites.

Because, in humans, the pathology of asthma affects mainly bronchi, any model of this disease should show differences in this compartment. In our study, a peribronchial infiltration of (activated) T cells was revealed with a substrain-specific difference. A peribronchial infiltration of T cells was also shown in human asthmatics [21], as well as in other animal models of allergic airway inflammations [22]. The depletion of CD4-positive T cells in a murine model completely prevented airway hyperreactivity (AHR) as well as infiltration of eosinophils [23], which supports a critical role of T cells in asthma. Also in the present study, AHR was found to be elevated (data not shown), as well as IgE levels and numbers of eosinophils, supporting the reliability of our model. However, the potential mechanisms for the substrain-specific recruitment of T cells to the bronchi remain to be revealed.

We considered either nervous system mediated effects or direct effects mediated via CD26. With regard to the first hypothesis, at least a direct interaction of nervous system-derived factors and T cell areas around the bronchi appears to be unlikely, as staining for p75NTR-positive structures and T cell areas revealed no overlap. Therefore, the expression of the multi-functional CD26 on the surface of the T cells alone or in combination with a different local chemokine pattern around the bronchi due to their local activity might lead to this differential increase. About 90% of all T cells in the lungs of wild-type rats were found to be CD26-positive using the optical disector method in the first series of experiments (data not shown). Activated T cells that are also distributed differentially in the airways express even more CD26 on their surface, which might potentiate ongoing processes.

First of all, CD26 is a potential mediator of adhesion to extracellular matrix proteins such as collagen and fibronectin [24,25]. This might retain more CD26-positive T cells in the lungs directly around the bronchi. Interaction of CD26 and fibronectin during adhesion and metastasis of rat breast cancer cells has been shown before [26], but its role under normal conditions has been examined to a lesser extent [27].

The dipeptidyl peptidase activity of CD26 or DP4-like functional homologues might also be a reason for a differential peribronchial increase of T cells, these sites having the first allergen-contact during inhalation of OVA. However – a priori – in the case of blunted degradation in the CD26-deficient animals, a more prominent recruitment to the sites of inflammation would be expected due to a longer half-life of those chemokines, being substrates of DP4, e.g. eotaxin/CCL11 [28]. Eotaxin attracts eosinophils via its receptor CCR3 and a strong eosinophilia in the lung should attract many T cells. More eosinophils are expected in CD26-deficient rat substrains in an asthma model, but in our model of an acute asthmatic response this was not the case [14].

Another potential reason for the differential increase of T cells is their chemotactic potential, which might be modulated by the expression of CD26 on their surface. MCP-1 is an example of a T cell attractant, and an OVA-induced pulmonary T cell accumulation is abolished in the absence of MCP-1-mediated signals in a mouse model [29]. MCP-1 has been shown to be increased in the BAL fluid of allergic asthmatic patients [30], and a potential association between the gene regulatory region of MCP-1 and asthma susceptibility has been suggested [31]. Expression of the MCP-1 receptor was shown on T cells highly expressing CD26, and only these cells responded to MCP-1 in chemotaxis assays [32]. In our study, qrtPCR revealed a significantly higher transcription of MCP-1 in the large airways of wild-type and CD26-deficient OVA-challenged rats compared to unchallenged rats. A higher expression of MCP-1 might lead to a peribronchial accumulation of T cells in these rats and the expression of CD26 on the surface of T cells in the wild-type rats might attract even more T cells in this substrain, according to results from other groups [32]. The differences in the lung parenchyma do not lead to a higher density of T cells in the CD26-deficient rats, which might be due to the lower transcription level in the parenchyma compared with the airways.

Another chemoattractant for T cells is SDF-1. It is cleaved by CD26 [7] and inactivated after its cleavage [33]. Within the airways of asthmatic patients, the immunoreactivity of SDF-1 is increased significantly and its expression is up-regulated in asthmatic tissues [34]. SDF-1 is also described as a substrate for the CD26/DP4-functional homologue dipeptidyl peptidase 8 (DP8) [35]. DP8 and its functional homologue DP9 are both expressed in the airways of both substrains and their expression is comparable in the lungs of wild-type and DP4-deficient animals [8]. Therefore, differences in the lungs appear to be independent of the expression of DP8 and DP9. PCR analyses revealed significant differences between wild-type and CD26-deficient rats under naive and challenged conditions, which might be due to a differential degradation of SDF-1 by CD26 and the consecutive up-regulation. CD26-positive T cells in the wild-type rats are likely to be attracted by SDF-1, which then is degraded by the recruited CD26-positive T cells, as it represents a substrate of CD26 in the bronchi of these rats. As CD26-deficient rats do not express CD26 on their T cells, these T cells might be attracted to a lesser extent. Along with this shift of the local substrate–peptidase balance, ‘compensatory’ up-regulation of SDF-1 transcription is abrogated in the CD26-deficient rats. Consequently, fewer T cells will be attracted, resulting in fewer T cell numbers around the bronchi of these rats. These mechanisms might represent a good part of the ‘anti-inflammatory’ effects in asthma being attributable to CD26-deficiency.

The transcription levels of four additional substrates of CD26 were analysed: RANTES, Mig, IP-10 and I-TAC. These chemokines are chemoattractors for T cells and are involved potentially in the pathogenesis of asthma [29,36,37]. The up-regulation of Mig and IP-10 after asthma-induction in both substrains support this hypothesis. However, in contrast to MCP-1 and SDF-1, the transcription of these chemokines is higher in the lung parenchyma compared to the bronchi. This suggests their involvement in an overall recruitment of T cells, which does not differentiate between different anatomical sites of the lung. Apparently, the substrain-specific difference in the transcription of Mig in the lung parenchyma does not exert an effect on the recruitment of T cells to this compartment of the lungs.

In conclusion, we suggest that the differential peribronchial T cell increase is mediated by the microenvironment at these sites. This needs further examination, in particular the quantification of local CD26 substrate levels that might mediate T cell recruitment, as well as the differentiation between their N-terminal truncated and full-length forms.


The authors have no conflict of interest to declare.


This study was supported by the German Research Foundation (SFB 587, project B11). We thank Susanne Kuhlmann, Susanne Fassbender, Andrea Herden, Katja Menge and Olga Skljar for skilful technical assistance, and Sheila Fryk for the correction of the English.