A Novel Role for CD4⁺ T Cells in the Control of Cachexia¹ Free

Cachexia, characterized by dramatic weight loss and muscle atrophy, is a consequence of a number of chronic disorders, including AIDS (1), aging (2, 3, 4), and type 1 diabetes (TID)⁴ (5). An association between immunodeficiency and cachexia (6, 7, 8, 9, 10) has led us to speculate that a deficiency in the immune system might play a direct role in the development of cachexia, and that correcting or inhibiting the immune deficiency might allow protection.

A deficiency in CD4⁺ T cells that express a low density of the cell surface marker CD44 (CD44^low) has been associated with aging (11, 12, 13) and with the development of spontaneous tumors (14, 15). In addition, we have recently shown that a subset of CD4⁺CD44^low cells, defined by their expression of the lowest density of CD44 (CD4⁺CD44^v.low), is depleted in diabetic mice at the onset of cachexia, but not in diabetic mice that are not cachexic (16). These data implicate the CD4⁺CD44^v.low T cell subset as a hypothetical candidate for modulating the development of cachexia. CD44 is one of the well-established cell surface markers used to distinguish Ag-inexperienced (naive) from Ag-experienced (memory) CD4⁺ T cells in the mouse. Thus, naive CD4⁺ T cells express CD44^low and a high density of CD62L (CD62L^high), whereas memory cells express CD44 at a high density (CD44^high) (17, 18). Naive cells also express a high density of CD45RB (CD45RB^high) (19, 20, 21). By these criteria, the CD4⁺CD44^v.low T cell subset, which is deficient in diabetic mice that are cachexic, is a naive CD4⁺ T cell.

Autoimmune destruction of pancreatic β cells in TID results in low insulin production and high blood glucose levels (BGL) (22, 23). Without insulin treatment, individuals with TID develop cachexia (5, 24). We have shown that wasting in the NOD mouse, the well-established mouse model for TID (25, 26), is also due to cachexia, with a dramatic loss of skeletal muscle weight, significant muscle protein loss, and activation of the ubiquitin proteasome pathway (16). In addition, muscle atrophy in the NOD mouse is associated with the presence of DNA fragmentation and a significant loss in muscle DNA content, suggesting the possibility that TID cachexia, like some models of cancer cachexia, involves apoptosis (27, 28, 29).

Using the NOD mouse model for TID cachexia, we find that infusion of highly purified CD4⁺CD44^v.low cells, but not CD4⁺ cells that are depleted of CD44^v.low cells, into prediabetic NOD mice significantly delays the onset of wasting and muscle atrophy, but no effect on the severity of diabetes was detected. CD4⁺CD44^v.low cells also inhibit muscle atrophy when infused into NOD mice that already have diabetes. That the mechanism of protection induced by CD4⁺CD44^v.low cells is independent of an effect on TID is further suggested by the finding that CD4⁺CD44^v.low cells also inhibit muscle atrophy, including muscle protein and DNA loss, in a C57BL/6 mouse strain model for cancer cachexia. In addition, infusion of CD4⁺CD44^v.low cells, but not CD4⁺ cells depleted of CD4⁺CD44^v.low cells, results in significant inhibition of Lewis lung carcinoma cell (LL2)-induced CD4⁺ T cell lymphopenia in the cancer cachexia model, suggesting that they might modulate cachexia by a mechanism that involves protection from CD4⁺ T cell lymphopenia. To our knowledge, this is the first report that indicates a role for CD4⁺ T cells in protecting the host from muscle atrophy, and suggests a novel role for a CD4⁺CD44^v.low cell subset, and the maintenance of immune homeostasis, in controlling cachexia.

NOD/LtJ (NOD) and C57BL/6J adult mice were purchased from The Jackson Laboratory. All protocols used in this study were conducted according to institutional guidelines and approved by the Institutional Animal Care and Use Committee.

Every 2 wk for the duration of the experiment, BGL were tested using a one-step Bayer Glucometer Elite (Bayer). Mice were considered diabetic when the BGL were >300 mg/dL over two consecutive readings.

LL2 is a cell line derived from the Lewis lung carcinoma. C57BL/6 mice were injected with 5 × 10⁵ LL2 in the left thigh. By day 7 post-LL2 injection, a small nodule can be detected in the thigh by palpitation. Mice were terminated and tissues were removed on or before day 28 post-LL2 injection, as indicated for each experiment.

NOD mice were weighed once per week for the duration of the experiment. C57BL/6 mice injected with LL2 cells were weighed once per week for the first 2 wk post-LL2 injection and then daily. Mice were considered wasting when their body weight was 20% less than at the beginning of the experiment. Weight loss in excess of 20% was associated with morbidity and mortality, and therefore, wasting mice were sacrificed and tissues were taken for analysis within 24 h of wasting assessment, or before, as indicated in each experiment.

Spleen cells from either 2- to 4-mo-old mice were prepared for single-cell suspensions. RBC were removed with lysing buffer (Sigma-Aldrich), and the remaining spleen cells were resuspended in PBS with 1% FBS (Serologicals). Splenocytes were labeled with an allophycocyanin-conjugated CD4-specific mAb and PE-conjugated CD44-specific mAb and CD4⁺CD44^v.low cells, and CD4⁺ cells depleted of CD4⁺CD44^v.low cells, were sorted under high speed on a FACSVantage SE with TurboSort (BD Immunocytometry Systems). The CD4⁺CD44^v.low population was defined as the CD4⁺ cells that stain the weakest for CD44, and was typically 3–5% of the total CD4⁺ cells in nondiabetic NOD mice (16) and 1–2% of the total CD4⁺ cells in untreated C57BL/6 mice. To avoid contamination with CD4⁺CD44^int cells, only the weakest staining 2% of CD4⁺CD44^v.low cells in NOD and 0.8% in C57BL/6, and the brightest staining 80% of CD4⁺ cells depleted of CD4⁺CD44^v.low cells were collected for the experiments described in this study. All cell populations were sampled and analyzed using a FACSCalibur to confirm the purity of the sorted populations. Sorted CD4⁺ cell populations were greater than 98% CD4⁺ (data not shown). Cells were washed once in PBS after sorting and before i.p. injection into syngeneic recipients.

Mice were sacrificed at the indicated times, and the pancreas was removed and immediately placed in 10% neutral buffered formalin to be fixed. After 24 h, the pancreata were embedded in paraffin, and 4-μm sections were cut. Sections were tested for the presence of insulin-producing β cells by immunohistochemistry (30). Sections were stained for insulin with guinea pig anti-porcine insulin Ab (DakoCytomation) using the indirect immunoperoxidase method. Sections were hydrated and blocked with 10% normal goat serum (Vector Laboratories). The sections were then incubated overnight in primary Ab and treated with a biotinylated secondary Ab (Vector Laboratories), followed by an avidin/biotinylated enzyme complex (Vector Laboratories). Slides were incubated in the dark with the enzyme substrate, 0.05% 3,3′-diaminobenzidine (Sigma-Aldrich) and counterstained, dehydrated, and mounted with Permount (Fisher Scientific). Insulin-positive areas were stained brown. The insulin-positive area was accurately scored in pixels by measuring the brown coloration (insulin-positive area after insulin-specific staining) in each section of each pancreas using a Zeiss Axiovision camera and acquisition software (Carl Zeiss) and KS300 analyzing system (Carl Zeiss). For each pancreas, the islet area from three sections 20 μm apart was analyzed and the mean was calculated.

The left and right anterior and lateral thigh muscles were isolated, weighed, then individually wrapped in autoclaved aluminum foil and stored at −80°C until analyzed. The packed muscle was immersed in liquid nitrogen and ground with mortar and pestle. The powdered tissue was transferred into 1 ml of ice-cold homogenization buffer (0.01 M Tris, 2 mM EDTA, 0.15 M NaCl, 0.012 M Brij 96, 2.22 mM Nonidet P-40, 0.025 mM leupeptin, 0.025 mM aprotinin, and 0.025 mM 4-(2-aminoethyl)benzenesulphonyl fluoride hydrochloride) and homogenized with an electronic pellet pestle. The homogenates were incubated for 30 min at 4°C, and centrifuged at 14,000 × g for 10 min at 4°C. Supernatants were thawed and diluted 1/800 in distilled H₂O on ice. Soluble protein concentration was determined by mixing 160 ml of the diluted sample with 40 ml of Bio-Rad dye reagent (Bio-Rad) in a 96-well plate using BSA as the protein standard. Using this information, the weight of soluble protein in each whole muscle was calculated. Supernatant measurements were performed at least in duplicate. The plates were incubated for 10 min at room temperature and read at 595 nm on a microplate reader (Molecular Devices).

The lateral and anterior thigh muscles were excised from both hind legs of each mouse and weighed. Tissue samples (50 mg) were minced and then lysed in a 6 M guanidinium chloride buffer containing proteinase K (40 μg/ml) at 55°C for 2–4 h, and then treated briefly with DNase-free RNase following the DNeasy protocol (Qiagen). Aliquots of DNA were diluted in 1 M urea for total DNA concentration measurements using a fluorometric DNA assay (31) with Hoechst dye 33258 (Bio-Rad), and the weight of DNA in each whole muscle was calculated.

Single cell-suspensions of lymph nodes (cervical, mesenteric, inguinal, para-aortic) were labeled with an allophycocyanin-conjugated CD4-specific mAb (RM4-5) and PE-conjugated CD44-specific mAb (IM7). Allophycocyanin-conjugated rat IgG2a and PE-conjugated rat IgG2b were used as isotype controls. All cell populations were sampled and analyzed using a FACSCalibur with CellQuest version 3.3 software (BD Immunocytometry Systems), and the percentage and total number of CD4⁺CD44^v.low (lowest fluorescent intensity peak for CD44 expression), CD4⁺CD44^int (middle fluorescent intensity peak), and CD4⁺CD44^high (highest fluorescent intensity peak), as described previously (16), were determined. All mAbs and isotype controls were purchased from BD Pharmingen.

The significance of the effect of CD4⁺CD44^v.low cells on protection from cachexia on transfer into NOD recipients was assessed using the log rank (Mantel-Cox) test. The significance of the effect of CD4⁺CD44^v.low cells on insulin-secreting cells in the pancreas, the effect of LL2 cell treatment on CD4⁺ T cell lymphopenia, and the effect of CD4⁺CD44^v.low cells on inhibition of CD4⁺ T cell subset lymphopenia was determined using the Mann-Whitney test. Skeletal muscle weight loss, skeletal muscle protein, and DNA content in cachexic animals were determined using Student’s t test. A p value equal to or less than 0.05 is considered significant for all tests.

Fifteen NOD female mice were monitored for diabetes and wasting from the age of 10 wk. Ten of the mice were diabetic by 28 wk of age (Fig. 1,a), and seven of these diabetic mice were wasting by 7 wk postdiabetes onset (Fig. 1 b). As we have previously shown (16), none of the five nondiabetic mice became wasting during the time course of the experiment, and the diabetic mice that were wasting did not lose weight until after the onset of diabetes.

We have previously shown that wasting in the NOD mouse is due to cachexia, and that the onset of cachexia in the diabetic NOD mouse is associated with a significant loss of CD4⁺CD44^v.low cells in the spleen and lymph nodes (16). To test the hypothesis that diabetic mice are protected from cachexia by the presence of CD4⁺CD44^v.low cells, we first tested the ability of these cells to inhibit the onset of diabetes and wasting in the NOD mouse. Prediabetic female NOD mice were infused with either highly purified CD4⁺CD44^v.low cells or no cells, and then monitored for diabetes and wasting. CD4⁺CD44^v.low cells significantly decrease the rate of onset and the incidence of wasting compared with untreated NOD mice (Fig. 2,a; 67% wasting in untreated compared with 30% wasting in CD4⁺CD44^v.low cell treated; p = 0.05). In contrast, the infusion of CD4⁺CD44^v.low cells did not inhibit either the onset or incidence of diabetes (Fig. 2,b; 75% diabetic in untreated compared with 80% diabetic in CD4⁺CD44^v.low cell treated). When the total body weight of wasting and nonwasting diabetic mice (from Fig. 2,b) was compared (Fig. 2,c) and calculated as a percentage of the weight of each mouse at 10 wk of age (Fig. 2,d), we found that CD4⁺CD44^v.low cells significantly inhibited weight loss in diabetic mice (p = 0.05 at 13 wk post cell infusion, and p = 0.03 at 15 wk post cell infusion for both body weight loss and percentage of body weight loss), suggesting that CD4⁺CD44^v.low cells decrease the severity as well as the incidence of cachexia (Fig. 2, c and d).

In a separate experiment (Fig. 3), we compared the effect of infusing CD4⁺CD44^v.low cells into mice that were already diabetic, and compared this effect with the infusion of CD4⁺ T cells that were depleted of the CD4⁺CD44^v.low cell subset. Although CD4⁺CD44^v.low cells once again inhibited wasting compared with untreated NOD mice (triangles vs squares in Fig. 3, p = 0.02), CD4⁺ cells depleted of CD44^v.low cells did not (circles). Taken together, our data suggest a relationship between the CD4⁺CD44^v.low cell subset and modulation of cachexia, but not TID.

To confirm that the inhibition of wasting in diabetic NOD mice by CD4⁺CD44^v.low was associated with an inhibition of muscle atrophy, the skeletal muscle was isolated from the CD4⁺CD44^v.low-treated and untreated diabetic mice (described in Fig. 2), 15 wk post cell infusion, and weighed (Table I). The weight of skeletal muscle from diabetic NOD mice that received CD4⁺CD44^v.low cells was significantly greater than from untreated diabetic NOD mice (179 ± 27 mg in treated compared with 119 ± 9 mg in untreated; p = 0.005). Cachexia is associated with preferential loss of skeletal muscle mass (lean tissue mass). That is, in cachexia, the weight of skeletal muscle as a percentage of total body weight is less than it is in noncachexic mice. To determine whether CD4⁺CD44^v.low cells protected skeletal muscle preferentially or whether their effect was equivalent in skeletal muscle and the rest of the body weight, skeletal muscle was calculated as a percentage of total body weight for each mouse, and any ability of CD4⁺CD44^v.low cells to preferentially protect skeletal muscle was determined. In untreated nondiabetic mice, the skeletal muscle analyzed makes up 1.45 ± 0.06% of the total body weight (Table I). In contrast, in untreated diabetic mice, the same skeletal muscle is reduced to 0.76 ± 0.02% of the total body weight. If the CD4⁺CD44^v.low cells inhibit skeletal muscle weight loss to the same extent as they protect wasting in the rest of the body, we would predict that percentage of total body weight that is skeletal muscle would not be significantly different from that seen in untreated mice (0.76 ± 0.02%). However, if CD4⁺CD44^v.low cells preferentially protect skeletal muscle, we would expect skeletal muscle in treated mice to make up a significantly greater percentage of total body weight than in untreated mice (greater than 0.76 ± 0.02%). Data show that skeletal muscle weight in treated mice is 0.96 ± 0.12% of total body weight, and is significantly greater than that in untreated mice (0.96 ± 0.12% compared with 0.76 ± 0.02%, respectively, p = 0.03), suggesting that CD4⁺CD44^v.low cells preferentially inhibit muscle atrophy in diabetic NOD mice (Table I).

Although data shown in Fig. 2 indicate that CD4⁺CD44^{v low} cells do not affect the onset of diabetes, it does not exclude the possibility that they might inhibit the loss of insulin-secreting β cells to a level that is sufficient to modulate cachexia, but not diabetes. To address this issue, the effect of CD4⁺CD44^v.low cells on insulin-secreting β cell mass in the pancreas was tested. Prediabetic NOD mice were treated with and without CD4⁺CD44^v.low cells, and monitored for the development of diabetes and wasting. At 15 wk after cell infusion, the mice were sacrificed, and the pancreas was removed and assayed for the presence of insulin by immunohistochemistry. The relative amount of insulin in each pancreas was determined by measuring the insulin-positive area in pixels. The insulin area in the pancreas of diabetic mice that were untreated (7 of 9 mice were diabetic) was first compared with the insulin area in pancreas of diabetic mice that were treated with CD4⁺CD44^v.low cells (8 of 9 were diabetic; Fig. 4,a). No significant difference was seen in the number of pixels of insulin measured in the pancreata of diabetic untreated (Fig. 4,b) and diabetic treated (Fig. 4,c) mice. Whereas all of the diabetic mice in the untreated group were also wasting by this time point, only four of eight diabetic mice in the treated group were wasting. However, although there was a trend that suggested an increase in insulin in the nonwasting group (Fig. 4,d) compared with the wasting group (Fig. 4,e), no significant difference was seen between these two groups. In addition, there is no difference (Fig. 4,g) in the amount of insulin in the nondiabetic untreated (Fig. 4,h) and treated (Fig. 4,i) mice. The amount of insulin measured in pancreas from treated mice that were diabetic, but not wasting, was, at best, 1% of that measured in pancreas of nondiabetic treated and untreated mice (Fig. 4, g compared with d, respectively).

Although the data to date suggest that the CD4⁺CD44^v.low cells inhibit cachexia by a mechanism that is independent of any effect on diabetes, we tested this possibility further using a model of cachexia that does not require diabetes as its primary disease. The LL2 model was chosen from a number of models of cancer cachexia because, unlike other models of cancer cachexia, the mechanism of cachexia induced by LL2, like that in the NOD mouse, involves skeletal muscle apoptosis in addition to activation of the ubiquitin proteasome pathway.

To determine whether cachexia in the LL2 model of cancer cachexia is also associated with lymphopenia, C57BL/6 mice were injected with LL2 cells and sacrificed at various times between 15 and 28 days afterward. Skeletal muscle was isolated and weighed. Consistent with published data, significant weight loss was seen in the skeletal muscle of LL2-treated mice by 25 days after injection compared with age-matched control mice (Fig. 5,a). CD4⁺ T cell number in both spleens and lymph nodes was significantly reduced in cachexic mice compared with untreated control mice, indicating that the LL2 model for cancer cachexia is associated with CD4⁺ T cell lymphopenia (p = 0.02 for spleen and p = 0.04 for lymph node) (Fig. 5 b).

Having established the time of onset of cachexia in the LL2 model in our hands, we set out to determine whether infusion of CD4⁺CD44^v.low cells was also able to modulate cancer cachexia, as it does for diabetes-induced cachexia. C57BL/6 mice were injected with LL2, and on the same day infused with either CD4⁺CD44^v.low cells, CD4⁺ cells depleted of CD44^v.low cells, or no cells (Fig. 6). Skeletal muscle was removed and weighed on day 28 post-LL2 injection. Skeletal muscle weight was significantly greater in CD4⁺CD44^v.low cell-treated mice, compared with mice treated with either CD4⁺ cells that were depleted of CD44^v.low cells, or with no cells.

Similar data were obtained when CD4⁺CD44^v.low cells were infused 24 and 25 days post-LL2 injection, at a time when cachexia was clearly evident. Again, the severity of muscle atrophy was significantly reduced in LL2-treated mice that received the CD4⁺CD44^v.low cell infusion compared with mice that did not (Fig. 6).

Muscle atrophy in cachexia is associated with a dramatic loss of muscle protein. In addition, the cancer cachexia model used in this study is also associated with a significant loss in muscle DNA (28, 29, 30). To determine whether protection from muscle atrophy by infused CD4⁺CD44^v.low cells resulted in protection from protein and DNA loss, C57BL/6 mice were injected with LL2 and, on the same day, infused with either CD4⁺CD44^v.low cells, CD4⁺ cells depleted of CD44^v.low cells, or no cells. Skeletal muscle taken from mice treated with CD4⁺CD44^v.low cells 28 days after LL2 injection contained significantly more soluble protein (Fig. 7,a) and DNA (Fig. 7 b) than skeletal muscle taken from mice treated with either CD4⁺ cells depleted of CD44^v.low cells, or no cells.

C57BL/6 mice injected with LL2 and infused with either CD4⁺CD44^v.low cells, CD4⁺ cells depleted of CD44^v.low cells, or no cells were sacrificed 28 days later, and lymph nodes were removed. LL2-treated mice that were infused with CD4⁺CD44^v.low cells, but not CD4⁺ cells depleted of CD44^v.low cells, had significantly greater numbers of CD4⁺CD44^v.low (Fig. 8,a), CD4⁺CD44^int (Fig. 8,b), and CD4⁺CD44^high (Fig. 8,c) cells compared with mice that did not receive a CD4⁺ T cell infusion. The effect of CD4⁺CD44^v.low cell infusion on CD4⁺ T cell lymphopenia is similar in the spleen to that seen for lymph node (data not shown). However, data are shown for lymph node rather than spleen because lymphopenia in the lymph node is more dramatic than that seen in the spleen (Fig. 5 b). The data show that treatment with CD4⁺CD44^v.low cells can inhibit cancer-induced CD4⁺ T cell lymphopenia, and are consistent with the hypothesis that the mechanism that CD4⁺CD44^v.low cells use to modulate cachexia involves inhibition of CD4⁺ T cell lymphopenia.

Cachexia (1, 2, 3, 4, 5) is the term used to describe the overall severe weight loss, muscle wasting, and anorexia seen in patients with a variety of primary disorders, including cancer (32) and AIDS (1); certain autoimmune conditions, including TID (5), chronic infection (33), and sepsis (34); and is often present in aging individuals with failure to thrive syndrome (2, 3, 4, 7, 8). Muscle wasting specifically refers to the loss of muscle mass, preferentially in skeletal muscle. It has important clinical consequences, including impaired rehabilitation and shortness of breath (34). Although a number of treatments have been used with some success (35, 36, 37), further improvements in therapeutic approaches are needed. In this study, we focus on the role of CD4⁺ T cells in the control of muscle wasting in animal models of TID and cancer cachexia.

Cachexia is often associated with lymphopenia, and cachexia in the two primary disease models used in this study is no exception. The presence of lymphopenia in patients with cachexia is associated with decreased responsiveness to therapy and poor prognosis. Whether lymphopenia is a cause or an effect of cachexia is not yet known. Under healthy conditions, a balance between naive and memory T cell numbers (38, 39, 40) and the size of the T cell pool (41, 42, 43) are maintained at a constant level by homeostatic mechanisms. We have previously shown that lymphopenia in TID is associated with a preferential loss of CD4⁺CD44^v.low cells, but not CD4⁺CD44^low and CD4⁺CD44^high cells, at the onset of cachexia (16). Based on these data, we hypothesized that CD4⁺CD44^v.low cells promote protection from cachexia and that cachexia ensues, at least in part, as a result of their loss. This hypothesis predicts that infusion of CD4⁺CD44^v.low cells into cachexic or precachexic mice might protect from cachexia. Our data show that the transfer of highly purified CD4⁺CD44^v.low cells into prediabetic NOD mice significantly inhibits total body weight loss and skeletal muscle atrophy. In contrast, these cells have no effect on the rate of onset and incidence of TID. Moreover, modulation of cachexia was also seen in mice that were infused with CD4⁺CD44^v.low cells, but not CD4⁺ cells that are depleted of CD44^v.low cells, after the onset of diabetes. These data suggest that CD4⁺CD44^v.low cells protect from cachexia, but not autoimmune diabetes. However, treatment of TID patients with insulin can inhibit muscle protein breakdown (44, 45) by inhibiting proteolysis and ubiquitin-mediated proteasomal activity (46). Therefore, it was possible that CD4⁺CD44^v.low cells inhibited the autoimmune destruction of insulin-secreting β cells to an extent that was insufficient to prevent TID, but sufficient to protect from cachexia. Analysis of the pancreata from diabetic treated and untreated mice failed to show a significant effect of CD4⁺CD44^v.low cells on insulin-secreting β cell mass in diabetic NOD mice. These data strongly suggest that these CD4⁺CD44^v.low cells modulate cachexia by a mechanism that is independent of an effect on autoimmune diabetes. Moreover, CD4⁺CD44^v.low cells also inhibit wasting and muscle atrophy in a model of cancer cachexia in a nonautoimmune-susceptible mouse strain, further supporting the conclusion that these cells protect from cachexia by a mechanism that is independent of an effect on autoimmunity and insulin secretion. Taken as a whole, the data also suggest that the pathways that lead to cachexia in multiple primary disease states are, at least in part, overlapping. This is consistent with findings that show a common program of changes in gene expression in atrophied skeletal muscle isolated from rats with cancer cachexia and rats with chemically induced diabetes (47). It is important to note that these data do not confirm that the loss of this cell subset is the cause of cachexia in TID and cancer. However, the data do show that CD4⁺CD44^v.low cells are able to protect from cachexia when infused into cachexia or precachexic mice.

Lymphopenia can lead to organ-specific autoimmunity (48, 49), and a resolution of the lymphopenia can result in a resolution of autoimmune pathology (50). However, because skeletal muscle has not been described as a target for autoimmunity, it is unlikely that modulation of cachexia by CD4⁺CD44^v.low cells involves mechanisms that are relevant to protection from autoimmunity that is specific for skeletal muscle. Additional support for the conclusion that the mechanism of protection exerted by CD4⁺CD44^v.low cells on cachexia does not involve protection from autoimmunity comes from the analysis of the phenotype of the CD4⁺CD44^v.low cell population (16). Thus, CD4⁺CD44^v.low cells do not express CD25 (51, 52, 53), CD38 (54), and CD45RB^low (55, 56), cell surface markers that define regulatory cell subsets known to inhibit autoimmunity. In addition, the regulatory cell marker Foxp3 (57, 58) is also not expressed by CD4⁺CD44^v.low cells (J. Davies, unpublished observation). Naive CD4⁺ T cells that do not express regulatory cell markers have been shown to inhibit wasting caused by autoimmune colitis, but not wasting caused by cachexia (50).

CD4⁺ T cells expressing the lowest two densities of CD44 (denoted as CD44^v.low and CD44^int in this study) are generally considered naive CD4⁺ T cells (CD44^low), whereas cells that express a high density of CD44 (CD44^high) are generally considered memory CD4⁺ T cells (17, 18). Therefore, based on the density of expression of CD44, CD4⁺CD44^v.low cells are naive CD4⁺ T cells. CD4⁺CD44^v.low cells also express a high density of CD62L and intermediate/high density of CD45RB (16), further suggesting their naive status (19, 20, 21). In addition, whereas activated CD4⁺ T cells express CD25 (59) and CD38 (60), the CD4⁺CD44^v.low cell subset expresses neither CD25 nor CD38 (16), suggesting this cell subset is a resting naive CD4⁺ T cell. However, our data do not exclude the possibility that this cell subset is not functional in its naive, unactivated state, and that the CD4⁺CD44^v.low T cell subset might only delay cachexia after activation and/or differentiation into a memory cell subset.

In addition to modulating cachexia, CD4⁺CD44^v.low cell infusion significantly reduced the extent of CD4⁺ cell lymphopenia in the CD44^v.low, CD44^int, and the CD44^high subsets, protection of CD4⁺CD44^high cell numbers being the most significant. Inhibition of lymphopenia by CD4⁺CD44^v.low cells might result from differentiation and proliferation to CD4⁺CD44^int and CD4⁺CD44^high cells, as well as proliferation to maintain the CD4⁺CD44^v.low cell population itself. Alternatively, CD4⁺CD44^v.low cells might inhibit the depletion of CD4⁺ T cell subsets. These data strengthen the association between lymphopenia and cachexia by showing, for the first time, that protection from lymphopenia is associated with protection from cachexia. It is tempting to speculate that the mechanism used by CD4⁺CD44^v.low cells to protect from cachexia involves protection from CD4⁺ T cell lymphopenia, and in particular, protection of the memory cell pool. Alternatively, protection from cachexia might not involve protection from lymphopenia. Nevertheless, the finding that CD4⁺CD44^v.low cells protect from lymphopenia might have important clinical implications in improving responsiveness to therapy, irrespective of whether protection from lymphopenia plays a role in protection from cachexia.

The highly consistent balance demonstrated between T cell subsets in healthy individuals might suggest that such control over T cell subset numbers is necessary to maintain the health of the host. In an attempt to repopulate the depleted lymphocyte pool, lymphopenia is generally followed by proliferation of the remaining memory cell pool (61), and proliferation and differentiation of naive CD4⁺ cells to become memory CD4⁺ cells (62, 63) by homeostatic expansion. We suggest that, particularly under conditions in which the cause of lymphopenia is not removed, as is seen for memory CD4⁺ T cell loss at the onset of TID (16), homeostatic repopulation might not be equivalent for all immune cell subsets. This might be particularly detrimental if memory cells that secrete procachexic cytokines, such as IFN-γ (64, 65), IL-1 (66, 67, 68), IL-6 (66, 67, 68, 69, 70), TNF-α (71), and TGF-β (72, 73), are preferentially activated and expanded. Alternatively, CD4⁺CD44^v.low cells might inhibit cachexia by promoting the expansion of memory cells that secrete IL-4 (74) and IL-10 (75), cytokines known for their anticachexic properties. However, whether cytokine manipulation is a viable treatment strategy for cachexia is currently under debate (76, 77, 78). It is important to note that T cells are not required for the onset of cachexia, and that cancer cachexia can be induced with similar kinetics in immunodeficient and congenic immunocompetent BALB/c recipients (67). However, this does not exclude the possibility that, if the immune system is present, it might play a role in enhancing the effects of cachexia. It is also possible that CD4⁺CD44^v.low cells can protect from cachexia in immunodeficient recipients by a mechanism that does not involve inhibiting homeostatic expansion of memory cells.

To our knowledge, this is the first report of an immune cell subset that promotes protection from cachexia, and provides a new approach in the search for novel therapeutics for the treatment of this syndrome. Whether the ability of CD4⁺CD44^v.low cells to protect from cachexia reflects a novel function for CD4⁺ T cells or whether it reflects an established function that had not previously been linked specifically to CD4⁺ cells and cachexia has yet to be determined. Nevertheless, these findings provide a new insight into understanding the pathways that control the development of cachexia, and suggest a novel role for the immune system in maintaining skeletal muscle integrity.

The authors have no financial conflict of interest.