Daily Archives: November 27, 2011

Sensory Neurons Control in Diabetes

Vintage Ad #1,278: Hospital for Sick Children ...

Image by jbcurio via Flickr

TRPV1+ Sensory Neurons Control β Cell Stress and Islet Inflammation in Autoimmune Diabetes

Rozita Razavi1247, Yin Chan1247, F. Nikoo Afifiyan1247, Xue Jun Liu137, Xiang Wan13, Jason Yantha124, Hubert Tsui124, Lan Tang124, Sue Tsai5, Pere Santamaria5, John P. Driver5, David Serreze6, Michael W. Salter13, H.-Michael Dosch124

2 Department of Pediatrics, The Hospital for Sick Children, Research Institute, University of Toronto, Toronto, ON, Canada, M5G 1X8
3 Department of Physiology, The Hospital for Sick Children, Research Institute, University of Toronto, Toronto, ON, Canada, M5G 1X8
4 Department of Immunology, The Hospital for Sick Children, Research Institute, University of Toronto, Toronto, ON, Canada, M5G 1X8
5 Julia McFarlane Diabetes Research Centre and Department of Microbiology and Infectious Diseases, Faculty of Medicine, University of Calgary, Calgary, AB, Canada T2N 4N1
6 The Jackson Laboratory, Bar Harbor, ME 04609, USA

Received 17 May 2006; revised 22 August 2006; Accepted 3 October 2006. Published: December 14, 2006. Available online 14 December 2006.

Referred to by: Sensory Neurons Link the Nervous System and Autoimmune Diabetes
CellVolume 127, Issue 615 December 2006Pages 1097-1099,
Helene Bour-Jordan, Jeffrey A. Bluestone

Click Here!


In type 1 diabetes, T cell-mediated death of pancreatic β cells produces insulin deficiency. However, what attracts or restricts broadly autoreactive lymphocyte pools to the pancreas remains unclear. We report that TRPV1+ pancreatic sensory neurons control islet inflammation and insulin resistance. Eliminating these neurons in diabetes-prone NOD mice prevents insulitis and diabetes, despite systemic persistence of pathogenic T cell pools. Insulin resistance and β cell stress of prediabetic NOD mice are prevented when TRPV1+ neurons are eliminated. TRPV1NOD, localized to theIdd4.1 diabetes-risk locus, is a hypofunctional mutant, mediating depressed neurogenic inflammation. Delivering the neuropeptide substance P by intra-arterial injection into the NOD pancreas reverses abnormal insulin resistance, insulitis, and diabetes for weeks. Concordantly, insulin sensitivity is enhanced in trpv1−/− mice, whereas insulitis/diabetes-resistant NODxB6Idd4-congenic mice, carrying wild-type TRPV1, show restored TRPV1 function and insulin sensitivity. Our data uncover a fundamental role for insulin-responsive TRPV1+ sensory neurons in β cell function and diabetes pathoetiology.


Type 1 diabetes (T1D) is an autoimmune disease governed by multiple genetic and environmental risk factors. Overt diabetes reflects glucose intolerance due to insulin deficiency. It is the end result of prediabetes, with progressive lymphoid infiltration around and then inside pancreatic islets of Langerhans and subsequent destruction of insulin-producing β-cells by autoreactive T lymphocytes (Anderson and Bluestone, 2005). T1D is characterized by a permissive immune system that fails to impose tolerance to arrays of self-antigens. Although the initiating events are not fully understood, β cell stress and death in the course of early islet restructuring are thought to provide sensitizing autoantigens, which expand autoreactive T cell pools in pancreatic lymph nodes ( [Mathis et al., 2001] [Rosmalen et al., 2002] [Trudeau et al., 2000] and [Zhang et al., 2002] ).

Self-antigens targeted in T1D are expressed in β cells and, in most cases, elsewhere in the body. They prominently include neuronal antigens, recognized by T cells with pathogenic potential ( [Salomon et al., 2001] and [Winer et al., 2001] ). It is unclear why, in T1D, T cells infiltrate only islets and their associated glia (Winer et al., 2003). It is also unclear whether autoimmunity and islet inflammation are related to hyperinsulinism and insulin resistance typical for even young NOD mice ( [Amrani et al., 1998] and [Chaparro et al., 2006] ).

There is evidence for functional interactions between nervous and immune systems (e.g., Wang et al. [2003]), but connections between islet autoimmunity and the nervous system remain ill defined (Carrillo et al., 2005). The interface between nervous system and external and tissue environments is the primary sensory afferent neuron. Primary afferents also have efferent function through local release of mediators such as neuropeptides (e.g., substance P [sP], CGRP). Islets may be innervated by primary sensory neurons, but their local function is uncertain (Ahren, 2000). A prominent subset of sensory neurons expresses the transient receptor potential vanilloid-1 (TRPV1) protein, a nonspecific cation channel that was first identified as the receptor for capsaicin ( [Caterina et al., 2000] and [Prescott and Julius, 2003] ). TRPV1+neurons are of known importance in proinflammatory reactions (O’Connor et al., 2004), islet infiltrating lymphocytes express receptors for neuropeptides (Persson-Sjogren et al., 2005), and we asked if these sensory neurons may have a role in T1D.


TRPV1+ Sensory Afferents Control Onset of Islet Inflammation and Diabetes

Using immunofluorescence, we determined that murine islets are associated with meshworks of TRPV1+ fibers (Figure 1A). TRPV1 was undetectable in endocrine islet cells by immunofluorescence (Figures 1A and 1B) and by RT-PCR of purified NOD islets (Figure S1 in the Supplemental Data available with this article online). Based on this evidence of islet innervation by TRPV1+ primary afferent sensory neurons, we investigated their possible role in T1D pathogenesis, using neonatal treatment of diabetes-prone NOD mice with capsaicin to permanently remove these neurons ([Caterina and Julius, 2001] and [Jancso et al., 1977] ). Two-day-old NOD mice received capsaicin (50 mg/kg, s.c.) or vehicle (NODcaps, NODctrl). As expected from the voluminous literature, capsaicin-treated mice were viable, fertile, and without abnormalities in growth or gross tissue structure, including pancreas. We confirmed the lack of TRPV1 expression in NODcaps mice, using immunofluorescence (Figures 1A and 1B, TRPV1 green), western blots (8, 12, and 20 week), and standard hot-plate testing in randomly selected 5- to 6-week-old mice or in diabetic animals (Figure 1C). Consistent with loss of neuropeptide secreting TRPV1+ neurons, NODcaps mice showed no sP staining (Figure S2).

Figure 1.

Removal of TRPV1+ Neurons Reduces Islet Infiltration and Diabetes Progression

Immunohistochemistry of TRPV1+ neurons in pancreas of 8-week-old NODctrl (A) and NODcaps (B). Insulin, blue; GFAP, red; and TRPV1, green; insert, dual-color stain for TRPV1 and GFAP in an adjacent, serial section (A). Western blot analysis of TRPV1 expression in spinal cord protein extracts from NODctrl and NODcapsmice at 12 (first lane) and 20 weeks of age (second lane) (C). Insulitis was scored in at least 300 islets from NODctrl and from NODcaps mice, 20 weeks of age (n = 5/group) (D). Kinetics of insulitis (E) and diabetes development in NODctrl and NODcaps mice (F).

Islet infiltration by hematopoietic inflammatory cells begins by 4–5 weeks of age, accumulating at the peri-islet Schwann cell (pSC) border; the autoimmune destruction of the pSC mantle is extensive by 8 weeks of age, 8–12 weeks before the onset of overt diabetes (Winer et al., 2003). In NODcaps mice, islet infiltrations were significantly reduced, compared with NODctrl (Figure 1D, p < 0.0001). More than 70% of NODcaps islets were free of lymphocytes by the age of 20 weeks, whereas even in nondiabetic NODctrl mice of that age, very few islets were lymphocyte free (Figure 1D). In one-third of NODcaps mice, infiltrated islets were entirely absent, as demonstrated by extensive, serial sectioning. In the remaining two-thirds of NODcaps mice, most islets remained free of lymphocytes, with some degree of inflammation in rare islets, but strikingly, there was little of the typical insulitis progression over time (Figure 1E). Concordantly, capsaicin treatment delayed the onset of diabetes (p = 0.0002) and reduced its incidence (Figure 1F, p < 0.0001, 35 weeks, n = 52 mice/group, life table analysis). No further NODcaps mice developed disease over the next 12 weeks, with about 80% reduction in final diabetes incidence (p < 0.0001, Fisher’s exact test).

NOD mice spontaneously develop a Sjögren-like sialitis/lacrimitis, which is under genetic controls separate from diabetes ( [Boulard et al., 2002] and [Cha et al., 2002] ). NODcaps mice exhibited the same submandibular lymphocyte infiltrates as NODctrl(Figure S3). Capsaicin treatment thus causes a dramatic reduction in pancreatic islet inflammation and development of diabetes without a generalized effect on autoimmune infiltrations elsewhere in the NOD mouse.

Capsaicin treatment effects on islet infiltration and disease development could reflect a failure to generate islet autoreactive T cell pools, a block of their tissue accumulation, or changes in immunoregulatory mechanisms. Capsaicin was reported to affect some immune functions in other animal models ( [Chancellor-Freeland et al., 1995] [Helme et al., 1987] [Nilsson et al., 1991] and [Santoni et al., 1996] ), and we compared development and functions of systemic (Figures 2A–2E) and pancreatic T cells (Figures 2F–2I) in NODcaps and NODctrl mice. Systemic T cell pools autoreactive to islet- (Insulin, GAD65), pSC- (GFAP, S100β), and other disease-associated antigens (HSP60, BSA) (Winer et al., 2003) were indistinguishable in NODcaps and NODctrlspleen cells (Figure 2A). To probe the development of diabetogenic T cell pools in NODcaps mice, we measured the peripheral frequency of pathogenic CD8+ T cells that recognize residues 206–214 of the glucose-6-phosphatase catalytic subunit related protein (IGRP) and its homologous, high-avidity mimotope, NRP-V7 ( [Amrani et al., 2000] [Amrani et al., 2001] [Anderson et al., 1999] [Lieberman et al., 2003] ,[Verdaguer et al., 1996] and [Verdaguer et al., 1997] ). The size of the circulating NRP-V7-reactive CD8+ T cell pool was similar in NODcaps and NODctrl spleens (0.25% ± 0.1%, p = 0.69) (Figure 2D). Lymphoid organ cellularities and T cell subset distributions were also not different in NODcaps and NODctrl, comparing splenocytes, axillary lymph nodes, and thymus (Figure S4). Delayed-type hypersensitivity reactions developed normally in NODcaps mice (Figure S5), suggesting intact antigen presentation and effector cell generation ( [Cua et al., 1995] and [Morikawa et al., 1993] ).

Figure 2.

T Cells in NODcaps and NODctrl

Splenic T cell proliferative responses to T1D-associated antigens; comparable results were obtained at different ages (A). Pooled flow cytometry data of CD4+(FITC) and of CD8+ (PE) splenic T cells, gated on CD3+ (APC) cells (n = 9/group) (B). Adoptive transfer of insulitis to NOD.scid mice by 6- to 8-week-old NODcaps or NODctrl splenocytes (n = 10/group) (C). Tetramer staining of IGRP TCR-specific T cells among CD8+ splenocytes, 8 weeks of age (n = 5) (D). Incidence of cyclophosphamide-induced diabetes (n = 6/group) (E), CD8+ (F), activated CD8+CD69+ (G), and CD4+CD25+ (H), and FoxP3+CD25+CD4+ T cells in pancreatic lymph nodes (I). Error bars represent + one SD.

In contrast, pancreatic NODcaps lymph node tissue contained significantly reduced proportions and absolute numbers of CD8+ and of activated CD8+CD69+ effector T lymphocytes critical for islet destruction (DiLorenzo et al., 1998) (Figures 2F and 2G). As a hallmark of prediabetes progression, prediabetic NOD mice selectively lose CD4+CD25+ and Foxp3+ regulatory T cell subsets in pancreatic lymph node tissue ([Bluestone and Tang, 2005] and [Pop et al., 2005] ). However, NODcaps mice maintained their regulatory T cell compartment in pancreatic lymph nodes beyond 12–16 weeks of age (Figures 2H and 2I). Thus, there are significant differences in the pancreatic, local immune system of NODcaps and NODctrl mice, consistent with the absence of chronic progressive islet inflammation in these animals.

Conceivably, undetected abnormalities in the NODcaps immune system might have influenced T cell pathogenicity. NODcaps animals that did develop disease had severe insulitis, and spleen cells from these animals transferred T1D with normal kinetics to lymphocyte-free NOD.scid recipients not treated with capsaicin (Figure S6). We also compared the ability of splenocytes from randomly selected NODcaps and NODctrl to initiate insulitis in such NOD.scid mice. NODcaps and NODctrl splenocytes initiated insulitis equally (Figure 2C). Moreover, we analyzed BDC2.5 T cell receptor transgenic NOD mice treated with capsaicin (BDC2.5caps) (Ji et al., 1999). BDC2.5capssplenocytes transferred T1D to NODctrl mice (p > 0.1). Low-dose cyclophosphamide accelerates NOD diabetes by multiple mechanisms (Hadaya et al., 2005). Low-dose cyclophosphamide accelerated diabetes development in both NODcaps and NODctrl (p = ns, Figure 2E) and was associated with reversal of the regulatory T cell accumulation present previously in NODcaps versus NODctrl pancreatic lymph nodes (Figure S7). Thus, NODcaps mice retain the principal ability to generate diabetogenic T cell pools.

Collectively, our observations separate loss of self-tolerance from target tissue invasion as distinct elements of T1D pathogenesis, and they demonstrate that the NODcapsimmune system retains pathogenic potential. TRPV1+ sensory neurons thus appear critical for the immune-cell accumulation in the pancreas.

NOD trpv1 Is Polymorphic

The above findings identify an important role of TRPV1+ primary afferent neurons in the initiation and progression of islet inflammation. TRPV1 maps to the Idd4.1 NOD diabetes-risk sublocus (mouse chromosome 11) into an ∼0.3 cM interval downstream of D11Ndsl (Figure 3A) ( [Grattan et al., 2002] [Ivakine et al., 2005] and [McAleer et al., 1995] ). Congenic replacement of the NOD Idd4 locus with the homologous B6 genomic interval protects from insulitis and, consequently, diabetes, although splenocytes from these congenic animals transfer both insulitis and diabetes to NOD.scid mice (Grattan et al., 2002). The NOD Idd4 risk locus differs from the homologous genomic region in the insulitis- and diabetes-resistant NOR strain, which carries nearly 90% of the NOD genome, including most other T1D risk loci ( [Ivakine et al., 2005] and [Serreze et al., 1994] ).

Figure 3.

trpv1 Mutations in NOD

trpv1 is located on mChr11 at 44.1 cM in the NOD Idd4.1 locus between D11Nds1 and D11Mit38 (top). Translated protein sequence of TRPV1NOD showing amino acid exchanges P322 → A322, D734 → E734, and corresponding regions in other species (bottom).

We cloned and sequenced TRPV1 cDNA from NOD and NOR mouse dorsal root ganglia (DRG) and confirmed selected sequence regions in NOD and NOR genomic DNA. The NOR trpv1 was identical to the published wild-type (B6, DBA) sequence, but the NOD sequence has two in-frame base exchanges, leading to predicted P322 → A322 and D734 → E734 amino acid replacements (Figure 3). Both replacements fall into regions highly conserved among diverse species (Figure 3).

TRPV1NOD Is Dysfunctional

We investigated whether the sequence differences in TRPV1NOD might cause abnormalities of TRPV1 function. The innervation of skin by TRPV1+ sensory afferents allowed assessment of potential functional differences by whole-animal experiments in which these afferents were stimulated by subcutaneous capsaicin application (Figure 4). Before testing capsaicin, we found that there were no differences between NOD and NOR mice in basal withdrawal responses to heat stimulation of the paw or tail (Figure 4A), indicating no generalized alteration of basal nociception in NOD mice. In addition, the sensitization of heat-evoked withdrawal responses after subcutaneous capsaicin administration was not different in NOD versus NOR mice (Figure 4B). However, nociceptive behavioral responses (biting, licking) evoked by subcutaneous capsaicin were markedly depressed in NOD as compared with NOR mice (p < 0.05,Figure 4C). Similarly, the paw edema produced by capsaicin was significantly reduced in NOD mice (p < 0.01, Figure 4D), implying reduced neuropeptide secretion at the site of stimulation. The depressed NOD-acute nociceptive and neurogenic inflammatory responses were not due to ongoing autoimmune inflammation, because NOD.scidmice, which lack lymphocytes, were not different from NOD (Figures 4E and 4F). Thus, the TRPV1NOD sequence abnormality appears to produce dysfunction of TRPV1-mediated responses to capsaicin.

Figure 4.

Depressed TRPV1-Mediated Responses in NOD Mice

Basal paw withdrawal (P) and tail flick (T) latencies to heat stimuli for NOD and NOR mice (n = 10/group) (A) and the latencies after capsaicin injection (B). Capsaicin-evoked (0.1 μg/10 μl) biting/licking during the first 5 min after capsaicin injection in NOD and NOR mice (C) or NOD.scid and NOD mice (E). Paw volume changes 45 min after capsaicin injection in NOD and NOR mice (D) or NOD.scidand NOD mice (F). Error bars represent + one SEM.

To assess TRPV1 function more directly, we recorded capsaicin-evoked Ca2+responses in DRG neurons from NOD and NOR mice (Figure 5). The maximum NOD DRG Ca2+ response to capsaicin was significantly smaller than that of NOR DRG neurons (p < 0.01, Figures 5A–5C). In addition, the maximum capsaicin response was reduced and required 10-fold higher drug concentrations in NOD DRG neurons, compared with that in NOR (p < 0.05, Figure 5C). In contrast, KCl-evoked Ca2+responses of NOD and NOR DRG neurons were not different (Figure 5D), indicating that NOD mice do not exhibit a general abnormality in Ca2+ responsiveness. The most-direct readouts of TRPV1 function are stimulus-evoked current responses, and we found that capsaicin-evoked whole-cell currents were significantly smaller in DRG neurons from NOD mice as compared with NOR mice (Figure 5E).

Figure 5.

Decreased TRPV1 Function and Expression in NOD DRG Neurons

Capsaicin-evoked Ca2+ response in cultured DRG neurons from NOD (A) and NOR mice (B), representative of five neurons/group. Dose-response curve of capsaicin-evoked Ca2+ responses in NOD and NOR DRG neurons (n = 4/group) (C). KCl-evoked Ca2+ response in NOD and NOR DRG neurons (D). Capsaicin-evoked currents with whole-cell patch-clamp recording from DRG of NOD and NOR mice (n = 5/group, no difference in NOD versus NOR average DRG capacitance) (E). TRPV1 expression in DRG extracts from NOD and NOR mice (n = 6/group) (F). Error bars represent + one SEM.

Because of the depressed TRPV1 function, we measured TRPV1 protein expression in DRGs and found that the basal TRPV1 protein level in NOD mice was lower than that in NOR (Figure 5F). Thus, the depression of capsaicin-evoked Ca2+ and current responses in DRG neurons from NOD mice may in part reflect decreased steady-state expression levels of TRPV1NOD. The rightward shift in the capsaicin concentration-response relationship suggests that the functionality of the TRPV1NOD protein itself may also be reduced as compared with TRPV1wild-type. Collectively, we discovered functional abnormalities in nociceptive behavior, neuropeptide secretion, channel function, and expression that define TRPV1NOD as a hypofunctional mutant.

Localized Pancreatic sP Administration Reverses Islet Pathology

We reasoned that abnormal TRPV1 function might selectively lead to islet pathology if there was a local, disease-predisposing TRPV1 effect on β cell function and if that effect was removed in NODcaps mice. The insulin-rich islet milieu represents a unique environment for TRPV1+ nerve terminals, as they express insulin receptors and insulin sensitizes and lowers the activation threshold of TRPV1 channels (Van Buren et al., 2005). Based on the diminished capsaicin-evoked neurogenic inflammation in NOD mice, and the reduced TRPV1 expression and function, we hypothesized that release of neuropeptides from sensory neuron terminals may be depressed in these mice. We examined a prominent neuropeptide, sP (O’Connor et al., 2004), and found that DRG sP levels were elevated in NOD compared with NOR mice (Figure S8). NODctrl and NOD.scid pancreas shows accumulation of more sP in nerve endings than B6 mice (Figure S2). Overall, the enhanced accumulation of sP is consistent with reduced sP release in NOD mice.

If depressed sP release was critical for NOD islet pathology, then increasing pancreatic sP levels should affect the pathogenic process, and we injected sP via the pancreatic artery. Figure 6A demonstrates the selective delivery of intra-arterially (i.a.) injected Evans Blue dye to pancreatic and pancreatic lymph node tissue (insert). In prediabetic NODctrl animals 12–14 weeks of age, we found that within 2 days after i.a. sP injection (2 nmoles/kg), about 80% of all islets were free of T cell infiltration (Figures 6B and 6C) and there were only small, residual infiltrates in the remainder. In animals receiving pancreatic i.a. vehicle injection, only 6% of islets were lymphocyte free (p < 0.0001) (Figure 6D). Systemic (i.v.) injection of the same sP dose did not have similar effects (see below, Figure 6I).

Figure 6.

i.a. Pancreas Injection of sP Reverses Islet Pathology

i.a. Evans Blue dye injection selectively stains pancreas (arrow) and pancreatic lymph node (insert) (A). Representative, H&E-stained pancreas sections from prediabetic (12 week) NOD females 2 days after i.a. pancreas injection with saline (B) or sP (C); arrow, insulitis lesions, and scale bars, ∼10 μm. Insulitis scores of 250–300 islets from saline- or sP-treated NOD prediabetics (D). Blood glucose levels in newly diabetic mice i.a. injected with saline (black line) or sP; red line, responder mice; blue line, nonresponder mice; each line represents one mouse (E). Insulin sensitivity of prediabetic (green), diabetic (black, i.a. saline injected) animals, and apparent nonresponder mice after i.a. pancreas sP injection (blue line) (F); graph shows changes in blood glucose after i.p. injection of insulin (n = 4/group). Flow cytometric measurement of CD4 and NK1R expression in CD3+prediabetic NOD lymphocytes; NK1R expression is upregulated after in vitro ConA stimulation (1 mg/ml, bottom) (G). Purified CD4+ T cells from NOD β2mnullsplenocytes were costimulated with plate-bound anti-CD3 and anti-CD28 in the presence of various concentrations of sP(H); graph shows mean incorporation of H3-TdR. (I) Clonal expansion, in pancreatic lymph nodes, of CFSE-labeled, CD4+BDC2.5 T cell receptor transgenic (diabetogenic) T cells, 4 days after transfer into saline (blue line) or sP treated (red line), 12-week-old normoglycemic NOD females. TCR-Vβ4+CFSE+ populations from two to five separate experiments are shown; pooled data are presented as mean ± SEM. Error bars represent ± one SD.

Analogous observations were made after pancreatic i.a. injection of sP into newly diabetic NOD mice, 2–3 days after diagnosis. After sP administration, and without insulin therapy, over half of the i.a. injected diabetics normalized blood glucose levels (Figure 6E, red lines). In these fully responsive mice, fasting blood glucose returned to near normal levels rapidly and remained at these levels for 2–8 weeks. Raising pancreatic sP levels dramatically enhanced insulin sensitivity, suggesting that the elevated insulin resistance at diagnosis was normalized (Figure 6F). On average, mice that reversed diabetes had less-extreme hyperglycemia at the time of diagnosis than did the nonresponding mice, likely reflective of a larger residual β cell mass at the time of sP administration. However, even in mice that failed to reverse hyperglycemia (blue lines), i.a. sP caused a significant improvement of metabolic control, preventing the progressive loss of body weight typical of overtly diabetic NOD mice (Figure S10). This improvement corresponds to significantly (p < 0.0001) improved insulin sensitivity (Figure 6F, blue line), which enhances the effectiveness of a small remaining β cell mass at diabetes onset. In all vehicle-injected control animals, blood glucose rose progressively, body weights declined, and animals were sacrificed because of severe diabetes between days 12 and 16.

Abundant expression of the NK1R sP receptor has been reported for islet infiltrating lymphocytes (Persson-Sjogren et al., 2005), and therefore, one likely target for sP is activated pancreatic T cells. We detected NK1R expression on a portion of T cells from pancreatic lymph nodes (Figure 6G). Upon in vitro activation with concanavalin A (Con A), essentially all NOD splenic T cells expressed NK1R (Figure 6G, insert). To determine the functional effect of NK1R ligation, we tested the sP response of activated CD4+ NOD T cells in vitro. sP abrogated cell proliferation and survival in a dose-dependent fashion (Figure 6H).

To determine the in vivo effect of pancreatic i.a. sP injection on clonal T cell expansion in pancreatic lymph nodes, we used islet-reactive, BDC2.5 T cell receptor transgenic T cells after labeling with a fluorescent dye, CFSE (Ji et al., 1999). Cells were transferred into 12-week-old normoglycemic NOD females, which had received pancreatic i.a. or systemic i.v. sP (red lines) or vehicle injections 12–16 hr prior (Figure 6I). BDC2.5 T cells from pancreatic lymph nodes were analyzed by flow cytometry 4 days later. Injection of sP reduced cellularity and clonal expansion, measured by dye dilution (p = 0.003). Systemic (i.v.) injection of the same sP dose had no effect on expansion of BDC2.5 T cells in pancreatic lymph nodes, suggesting a pancreas tissue-conditioning effect of i.a. pancreas injection that lasts at least 12–16 hr. A third set of animals received BDC2.5 cells that were in vitro pretreated with sP or vehicle overnight. This in vitro pretreatment with sP reduced the ability of these cells to expand in pancreatic lymph nodes (Figure 6I, bottom, p = 0.0045). As equal numbers of viable cells were transferred, these observations imply that sP also has a T cell conditioning effect.

Taken together, the data suggest that reduced neuropeptide release by pancreatic TRPV1+ nerve terminals is a pathogenic event in NOD diabetes amenable to therapeutic correction.

TRPV1 Function and β Cell Stress

β cell stress has previously been suggested as an early element of T1D pathoetiology, and we asked if the hypofunctional TRPV1NOD is related to signs of β cell stress, hyperinsulinism, and abnormal glucose clearance observed even in young NOD mice ([Rosmalen et al., 2000] and [van de Wall et al., 2005] ). We compared measures of β cell function in untreated and in capsaicin-treated NOD.scid mice, and in C57/BL6J (“B6”) and trpv1−/− mice, the latter with a normal complement of sensory afferent neurons but absent TRPV1 expression (Caterina et al., 2000). NOD.scid mice were used to ascertain absence of lymphoid islet infiltrations in NOD experiments, and CD1 mice provided controls (Figure S9).

The high-normal serum glucose levels after i.p. glucose challenge in 10- to 12-week-old NOD.scidctrl mice were significantly reduced in NOD.scidcaps mice (Figure 7A, p = 0.004). The improved NOD.scidcaps glucose response was associated with significantly less insulin production (Figure 7A, insert), suggesting more-effective insulin action after removal of TRPV1+ sensory neurons.

Figure 7.

Glucose Tolerance and Insulin Sensitivity in NODcaps and TRPV1null B6 Mice

Changes in blood glucose (mmol/L) (A) and insulin levels (insert) after i.p. glucose challenge (2 g/kg) in fasted NODctrl, NODcaps mice, and wild-type or TRPV1nullB6 mice (n = 5/group) at 6 to 8 weeks old (B); insert, parallel changes in plasma insulin levels. Insulin sensitivity: changes in blood glucose (mmol/L) after i.p. injection of insulin (Humulin, 0.75 U/kg) in similar groups of mice (C and D). Error bars represent ± one SD.

B6 mice develop elevated insulin resistance and type 2 diabetes-like disease (Parekh et al., 1998), attributed to the functional deletion of nicotinamide transhydrogenase (Freeman et al., 2006). Consistently, we observed high blood glucose levels after standard i.p. glucose challenge (Figure 7B). B6.trpv1−/− mice showed a significantly improved glucose response, analogous to NODcaps mice, raising the possibility that TRPV1 may play a general role in β cell physiology.

To more directly assess if both data sets could reflect enhanced insulin sensitivity due to TRPV1 removal, we measured glucose clearance after a single insulin injection. Compared to their respective control animals, NODcaps and B6.trpv1−/− mice showed significantly enhanced and accelerated glucose clearance, which we interpreted as evidence for reduced insulin resistance due to the absence of TRPV1 in these two independent animal models. Similar outcomes in NODcaps and B6.trpv1−/− mice link the observed effects on β cell function to TRPV1 itself. Enhanced insulin resistance associated with TRPV1NOD constitutes a persistent β cell stress, likely worsening with progressive islet inflammation (Nielsen et al., 2004). TRPV1 and TRPV1+ sensory neurons impact insulin/glucose homeostasis in these models of type 1 and type 2 diabetes.

Congenic Replacement of NOD.Idd4

As a test of our conclusions that TRPV1 plays a fundamental role in islet inflammation and insulin homeostasis, we investigated NOD.B6.Idd4-congenic mice (Figure 8). These mice carry wild-type TRPV1 in this locus and are insulitis- and diabetes resistant despite the fact that their splenocytes transfer diabetes to NOD.scid mice (Grattan et al., 2002). Consistent with wild-type trpv1 function, we found that these congenics have normalized behavioral responses to cutaneous capsaicin injection (biting/licking,Figure 8A), neurogenic inflammation after paw injection with capsaicin (Figure 8B), and Ca2+ responses in capsaicin-stimulated DRG (Figure 8C). The glucose responses of weaned Idd4 congenics were comparable to those in NOD control mice; however, there was a significant reduction in glucose-induced insulin secretion, suggesting an absence of elevated insulin resistance in these animals (Figure 8D).

Figure 8.

Rescue of TRPV1 Function and Insulin Resistance in NOD.B6.Idd4-Congenic Mice

Capsaicin-evoked (0.1 μg/10 μl) biting/licking during the first 5 min after capsaicin injection in NOD and NOD.B6.Idd4 mice (n = 6/group) (A). Paw volume changes 45 min after capsaicin injection in NOD (n = 4) and NOD.B6.Idd4 (n = 6) mice (B). Capsaicin-evoked Ca2+ response in cultured DRG neurons from NOD (n = 4 neurons) and NOD.B6.Idd4 (n = 4 neurons) mice (C). Insulin levels after i.p. glucose challenge (2 g/kg) in fasted NOD and NOD.B6.Idd4 mice (n = 5/group) (D). Error bars represent ± one SEM.

These NOD congenic mice resemble NODcaps mice, as both are insulitis/diabetes protected, although their T cells transfer diabetes to NOD.scid recipients. TRPV1NODadds elevated insulin resistance as a new, strikingly diabetes-relevant phenotype to the NOD Idd4.1 risk locus, which is presently associated only with insulitis; transgenic rescue experiments will be required for formal proof that trpv1 is or is not the Idd4.1diabetes risk gene.


In the different, independent animal strains and experimental conditions analyzed, TRPV1 emerges as a central controller of both islet stress and T cell infiltration. Elimination of TRPV1+ neurons by capsaicin, transient functional normalization by acute local sP injection, or replacement with wild-type trpv1 in Idd4 congenics has the same, islet-specific outcome: normalized insulin sensitivity and abrogation of insulitis, despite unimpeded generation of autoreactive lymphocytes that can transfer disease to untreated NOD hosts. The most parsimonious explanation unifying these observations is a local feedback interaction between β cells and the primary sensory neurons innervating islets (Figure S11), with nerve terminals responding to local insulin with release of neuropeptides that sustain β cell physiology in an optimal range. Normally, this interaction is in balance, but in the NOD mouse, hypofunction of TRPV1 unbalances the feedback, with β cell stress due to hyperinsulinism, insulin resistance, and infiltration by autoreactive T cell pools independently generated in the NOD mouse. Removing TRPV1 neurons leads to elimination of the unbalanced, pathogenic interaction, whereas administering sP exogenously may renormalize the interaction transiently.

A direct neuropeptide effect on β cells has previously been reported, with deleterious outcomes at low concentrations but β cell augmenting effects at higher concentrations ( [Barakat et al., 1994] [Bretherton-Watt et al., 1992] and [Hermansen and Ahren, 1990] ). Our hypothesis, based on the foregoing results that in NOD mice suppressed neuropeptide secretion is a pathogenic event, was positively answered through two independent approaches: removal of TRPV1+ neurons and local i.a. pancreas injection with sP, both with similar results. Pancreas sP injection normalized all parameters tested: clearing of insulitis lesions, enhancement of insulin sensitivity, and consequent reversal of overt diabetes that lasted for weeks. The only other strategy to reverse NOD diabetes is toxic immunosuppression with anti-CD3 antibodies, now also in clinical trials with human diabetics (Keymeulen et al., 2005).

Collectively, our findings are inconsistent with the view that diabetes is due solely to immunological and endocrine abnormalities. Rather, our observations demonstrate that the nervous system, in particular TRPV1+ primary afferent neurons, has a critical role in diabetes pathoetiology. Analogous findings in NODcaps, NOD.Idd4 congenics, and trpv1 knockout mice add strength to our conclusions, as does an earlier report demonstrating that another TRPV1-dependent neuropeptide, CGRP, prevents diabetes when transgenically overexpressed in the islet (Khachatryan et al., 1997). Recently, we generated preliminary evidence for insulitis- and diabetes protection by transsection of sensory nerves innervating the pancreas (J.Y., Y.C., H.T., L.T., R.R., M.W.S., and H.-M.D., unpublished data), providing yet another line of support for the role of TRPV1+sensory neurons in T1D pathoetiology.

The mapping of several NOD disease-associated phenotypes to a single, mutant protein, TRPV1, implies that TRPV1+ sensory afferents are key elements for normal islet physiology, opening broad new areas of research, including insulin resistance (Moesgaard et al., 2005), which remains a challenge after decades of intense investigation (LeRoith and Gavrilova, 2006).

Our data allowed us to identify the molecular mechanism that translates a system-wide genetic trpv1 defect into pancreas-specific disease. TRPV1+ sensory neurons express high-affinity insulin receptors, and insulin potentiates TRPV1 currents (Van Buren et al., 2005) and lowers TRPV1 thermal activation thresholds (Sathianathan et al., 2003). At body temperature, the insulin-rich islet milieu should generate tonic TRPV1 current activation with associated neuropeptide release impacting on basal insulin secretion, a local control circuit first envisioned over a decade ago (Hermansen and Ahren, 1990). In NOD mice, this sensory nerve terminal-β cell circuit has gone astray, with disease prevention through either its removal or through local supply of sufficient neuropeptide.

We demonstrated that TRPV1+ sensory afferents control pancreatic tissue access for immune cells, which may occur through modifying their immigration, residence, emigration, or a combination of these elements. It is likely that progressive islet infiltration will also compound β cell stress, which we believe is central to T1D pathoetiology. There is human disease precedence for a role of sensory neurons controlling lymphocyte tissue access, because rare patients without sensory nerves (CIPA syndrome) succumb to massive infections with little tissue infiltration, despite normal in vitro immune functions (Indo et al., 1996).

We discovered mutations in the coding sequence of TRPV1NOD gene contained within the Idd4 diabetes risk locus ( [Grattan et al., 2002] [Ivakine et al., 2005] and [McAleer et al., 1995] ). NOD.B6.Idd4 congenic mice show normalized behavioral, electrophysiological, and insulin-resistance phenotypes. Intriguingly, the TRPV1 locus is contained within other overlapping autoimmune loci (eae7orch3, and streptozotocin sensitivity) ( [Babaya et al., 2005] [Butterfield et al., 1999] and [Butterfield et al., 1998]), raising the possibility that TRPV1 may play a role in other autoimmune conditions. Indeed, B6 mice, relatively resistant to streptozotocin-induced T1D, show increased streptozotocin susceptibility in B6.trpv1−/− mice (data not shown).

Collectively, our findings identify TRPV1+ sensory neurons as important elements of diabetes pathoetiology, with effects that provide rational mechanisms of the tissue selectivity of the disease, its links to β cell physiology, stress, and insulin resistance. Our observations open new avenues for therapeutic strategies, raising the possibility that sensory nerve dysfunction may contribute to prediabetes initiation and progression in diabetes-prone humans.

Experimental Procedures


All mice were obtained from the Jackson Laboratories (Bar Harbor, Maine) and maintained under approved protocols in our vivarium (NOD female T1D incidence: 85%–90%). In adoptive transfer experiments, splenocytes from three to five donors indicated were pooled and 107 viable cells/mouse were injected (100 μl i.v.) into irradiated (300 rad) recipients. Diabetes was defined as diabetic blood glucose measurements (>13.8 mM/l) on two consecutive days.

T Cell Studies

Splenocytes (4 × 105 cells/well) or lymph node cells (2 × 105 lymph node cells plus 2 × 105 irradiated [1100 rad], syngeneic splenocytes/well) from 3- to 24-week-old NOD females were cultured in AIM V serum-free medium containing T1D target antigens (0.002–50 μg/ml) as described (Winer et al., 2003). [3H]thymidine incorporation was measured after 72 hr. To normalize pooled data, we calculated a stimulation index (SI, cpm antigen stimulated/medium control). In some proliferation studies, plate-bound anti-CD3 (0.001–3 μg/m) and anti-CD28 (0.2 μg/ml) (BD PharMingen) were used to stimulate CD4+ T cells negatively selected from NOD β2mnull splenocytes.

i.a. Pancreas Injection

The aorta was developed with minimal trauma and ligated just prior to the celiac branching and injected (32G needle) with Evans Blue (3 mg/kg/100 μl; Sigma), sP (2 nmol/100 μl; Sigma), or saline. Ligations were released after closure of the injection site.

5- and 6-Carboxyl-Fluorescien Succimidyl Ester Labeling

For dye dilution in vivo clonal expansion studies, splenic CD4+ T cells from NOD-BDC2.5 females were incubated with 2.5 μM CFSE (10′/37°C, Molecular Probes, Eugene, OR) in PBS. Prediabetic (12 week) NOD females pretreated 12 hr prior with sP or saline were injected i.v. with 3 × 106 CFSE-labeled CD4+ T cells.

Immunofluorescence and Histology

Frozen murine pancreas sections were fixed in 4% paraformaldehyde, blocked, and stained with variously fluorescence- or biotin-conjugated antibodies or strepavidin as described (Winer et al., 2003). TRPV1 staining was performed on snap-frozen sections of NOD female pancreas with an overnight incubation of primary antibody at 4°C. To score insulitis severity, pancreata were fixed in 10% buffered formalin for a minimum of 24 hr. Histological sections were stained with hematoxylin and eosin and scored by three blinded observers (Winer et al., 2003). NRP-V7/H-2Kd tetramer positivity was analyzed on gated CD8+B220 cells and reported as percentage of cells binding the tetramer minus the percentage of cells binding the negative control tetramer TUM/H-2Kd.

Molecular Cloning

PCR amplification used TRPV1-specific primers with cDNA from NOD, NOR, and B6 DRG, cloning followed standard procedures using the TOPO XL PCR Cloning Kit (InVitrogen, Mississauga, ON).

Behavioral Studies and Paw Volume Measurement

All behavioral tests of the mouse strains indicated were conducted between 9 and 16 hr after appropriate acclimatization to the testing environment and with standard intrumentation/apparatus. Paw volume measurements used a plethysmometer (Ugo Basile), and values were standardized as a percentage of individual preinjection volumes.

Ca2+ Response Measurement

DRG from donors indicated were isolated and cultured 3–5 days in F12 medium with 10 ng/mL nerve- and 10 ng/mL glial-derived NGF. Fura-2 (Molecular Probes, Eugene, OR) was used to assess [Ca2+]i by ratiometric measurement. Each 340 nm image was divided, on a pixel-by-pixel basis, by the corresponding 380 nm image, producing a ratio. Averaged values of the ratios within each region of interest were plotted as a function of time. Whole-cell DRG patch-clamp recordings were performed at room temperature. TRPV1 currents were recorded using an Axopatch 1-D amplifier, data were digitized with DigiData1322, filtered (2 kHz), and acquired by the pClamp9.0 program. Recordings in which the series resistance varied by more than 10% were rejected.


DRG or dorsal horns (spinal cord) were dissected and snap-frozen. Extracted total protein (45 μg DRG protein, 30 μg spinal cord) was electrophoresed (10% acrylamide gels), western blotted, probed overnight with rabbit anti-TRPV1 antibody (1:250, Oncogene), and developed with the ECL kit (Amersham).


All tests were two tailed, significance was set at 5%. Life tables, t tests (flow cytometry), ANOVA, and Fisher’s exact test were used as described in the text.


This work was supported by Canadian Institute of Health Research grants (H.-M.D., P.S., and M.W.S.). P.S. is a Scientist of the Alberta Heritage Foundation for Medical Research. R.R. is recipient of a CIHR-Industry scholarship, Y.C. a recipient of Banting and Best and CIHR scholarships, X.W. of a fellowship from the Heart and Stroke Fdn, X.J.L. of a fellowship from the Can. Arthritis Network, and F.N.A. of a scholarship from the Can. MS Soc. We thank Drs D. Ize-Ludlow, D.J. Becker (Children’s Hospital of Pittsburgh), and J. Dupré (Robarts Institute, University of Western Ontario, London, ON) for helpful discussions.


Ahren, 2000 B. Ahren, Autonomic regulation of islet hormone secretion–implications for health and disease. Diabetologia 43  (2000), pp. 393–410. | View Record in Scopus | | Cited By in Scopus (33)

Amrani et al., 2000 A. Amrani, J. Verdaguer, P. Serra, S. Tafuro, R. Tan and P. Santamaria, Progression of autoimmune diabetes driven by avidity maturation of a T- cell population. Nature 406  (2000), pp. 739–742.

Amrani et al., 2001 A. Amrani, P. Serra, J. Yamanouchi, J.D. Trudeau, R. Tan, J.F. Elliott and P. Santamaria, Expansion of the antigenic repertoire of a single T cell receptor upon T cell activation. J. Immunol. 167  (2001), pp. 655–666. | View Record in Scopus | | Cited By in Scopus (107)

Anderson and Bluestone, 2005 M.S. Anderson and J.A. Bluestone, The NOD mouse: a model of immune dysregulation. Annu. Rev. Immunol. 23  (2005), pp. 447–485. |View Record in Scopus | | Cited By in Scopus (9)

Barakat et al., 1994 A. Barakat, G. Skoglund, C. Boissard, G. Rosselin and J.C. Marie, Calcitonin gene-related peptide and islet amyloid polypeptide stimulate insulin secretion in RINm5F cells through a common receptor coupled to a generation of cAMP. Biosci. Rep. 14  (1994), pp. 1–13. | View Record in Scopus | | Cited By in Scopus (132)

Boulard et al., 2002 O. Boulard, G. Fluteau, L. Eloy, D. Damotte, P. Bedossa and H.J. Garchon, Genetic analysis of autoimmune sialadenitis in nonobese diabetic mice: a major susceptibility region on chromosome 1. J. Immunol. 168  (2002), pp. 4192–4201. | View Record in Scopus | | Cited By in Scopus (10)

Butterfield et al., 1998 R.J. Butterfield, J.D. Sudweeks, E.P. Blankenhorn, R. Korngold, J.C. Marini, J.A. Todd, R.J. Roper and C. Teuscher, New genetic loci that control susceptibility and symptoms of experimental allergic encephalomyelitis in inbred mice.J. Immunol. 161  (1998), pp. 1860–1867. | View Record in Scopus | | Cited By in Scopus (77)

Carrillo et al., 2005 J. Carrillo, M.C. Puertas, A. Alba, R.M. Ampudia, X. Pastor, R. Planas, N. Riutort, N. Alonso, R. Pujol-Borrell and P. Santamaria, et al. Islet-infiltrating B-cells in nonobese diabetic mice predominantly target nervous system elements.Diabetes 54  (2005), pp. 69–77. | View Record in Scopus | | Cited By in Scopus (717)

Caterina et al., 2000 M.J. Caterina, A. Leffler, A.B. Malmberg, W.J. Martin, J. Trafton, K.R. Petersen-Zeitz, M. Koltzenburg, A.I. Basbaum and D. Julius, Impaired nociception and pain sensation in mice lacking the capsaicin receptor. Science 288  (2000), pp. 306–313. | Arthritis Rheum.,  46  (2002), pp. 1390–1398. | View Record in Scopus | | Cited By in Scopus (66)

Chaparro et al., 2006 R.J. Chaparro, Y. Konigshofer, G.F. Beilhack, J.A. Shizuru, H.O. McDevitt and Y.H. Chien, Nonobese diabetic mice express aspects of both type 1 and type 2 diabetes. Proc. Natl. Acad. Sci. USA 103  (2006), pp. 12475–12480. | View Record in Scopus | | Cited By in Scopus (52)

DiLorenzo et al., 1998 T.P. DiLorenzo, R.T. Graser, T. Ono, G.J. Christianson, H.D. Chapman, D.C. Roopenian, S.G. Nathenson and D.V. Serreze, Major histocompatibility complex class I-restricted T cells are required for all but the end stages of diabetes development in nonobese diabetic mice and use a prevalent T cell receptor a chain gene rearrangement. Proc. Natl. Acad. Sci. USA 95  (1998), pp. 12538–12543. | View Record in Scopus | | Cited By in Scopus (44)

Grattan et al., 2002 M. Grattan, Q.S. Mi, C. Meagher and T.L. Delovitch, Congenic mapping of the diabetogenic locus Idd4 to a 5.2-cM region of chromosome 11 in NOD mice: identification of two potential candidate subloci. Diabetes 51  (2002), pp. 215–223. | View Record in Scopus | | Cited By in Scopus (17)

Helme et al., 1987 R.D. Helme, A. Eglezos, G.W. Dandie, P.V. Andrews and R.L. Boyd, The effect of substance P on the regional lymph node antibody response to antigenic stimulation in capsaicin-pretreated rats. J. Immunol. 139  (1987), pp. 3470–3473. |View Record in Scopus | | Cited By in Scopus (14)

Indo et al., 1996 Y. Indo, M. Tsuruta, Y. Hayashida, M.A. Karim, K. Ohta, T. Kawano, H. Mitsubuchi, H. Tonoki, Y. Awaya and I. Matsuda, Mutations in the TRKA/NGF receptor gene in patients with congenital insensitivity to pain with anhidrosis. Nat. Genet. 13  (1996), pp. 485–488. | View Record in Scopus | | Cited By in Scopus (16)

Jancso et al., 1977 G. Jancso, E. Kiraly and A. Jancso-Gabor, Pharmacologically induced selective degeneration of chemosensitive primary sensory neurones. Nature, 270  (1977), pp. 741–743. | View Record in Scopus | | Cited By in Scopus (21)

Keymeulen et al., 2005 B. Keymeulen, E. Vandemeulebroucke, A.G. Ziegler, C. Mathieu, L. Kaufman, G. Hale, F. Gorus, M. Goldman, M. Walter and S. Candon, et al.Insulin needs after CD3-antibody therapy in new-onset type 1 diabetes. N. Engl. J. Med. 352  (2005), pp. 2598–2608. | View Record in Scopus | | Cited By in Scopus (32)

LeRoith and Gavrilova, 2006 D. LeRoith and O. Gavrilova, Mouse models created to study the pathophysiology of Type 2 diabetes. Int. J. Biochem. Cell Biol. 38  (2006), pp. 904–912. Article |  PDF (283 K) | | View Record in Scopus | | Cited By in Scopus (188)

Mathis et al., 2001 D. Mathis, L. Vence and C. Benoist, beta-Cell death during progression to diabetes. Nature 414  (2001), pp. 792–798. | View Record in Scopus | | Cited By in Scopus (104)

Moesgaard et al., 2005 S.G. Moesgaard, C.L. Brand, J. Sturis, B. Ahren, M. Wilken, J. Fleckner, R.D. Carr, O. Svendsen, A.J. Hansen and D.X. Gram, Sensory nerve inactivation by resiniferatoxin improves insulin sensitivity in male obese Zucker rats.Am. J. Physiol. Endocrinol. Metab. 288  (2005), pp. E1137–E1145. | Cell. Immunol., 152  (1993), pp. 200–210. Article |  PDF (382 K) | | View Record in Scopus | | Cited By in Scopus (13)

Nilsson et al., 1991 G. Nilsson, K. Alving and S. Ahlstedt, Effects on immune responses in rats after neuromanipulation with capsaicin. Int. J. Immunopharmacol. 13  (1991), pp. 21–26. Article |  PDF (499 K) | |  T.M. O’Connor, J. O’Connell, D.I. O’Brien, T. Goode, C.P. Bredin and F. Shanahan, The role of substance P in inflammatory disease. J. Cell. Physiol., 201  (2004), pp. 167–180. | View Record in Scopus | | Cited By in Scopus (70)

Persson-Sjogren et al., 2005 S. Persson-Sjogren, K. Lejon, D. Holmberg and S. Forsgren, Expression of the NK-1 receptor on islet cells and invading immune cells in the non-obese diabetic mouse. J. Autoimmun. 24  (2005), pp. 269–279. Article | PDF (933 K) | | View Record in Scopus | | Cited By in Scopus (130)

Prescott and Julius, 2003 E.D. Prescott and D. Julius, A modular PIP2 binding site as a determinant of capsaicin receptor sensitivity. Science 300  (2003), pp. 1284–1288. | View Record in Scopus | | Cited By in Scopus (39)

Rosmalen et al., 2002 J.G. Rosmalen, P.J. Leenen, C. Pelegri, H.A. Drexhage and F. Homo-Delarche, Islet abnormalities in the pathogenesis of autoimmune diabetes.Trends Endocrinol. Metab. 13  (2002), pp. 209–214. Article |  PDF (54 K) | | View Record in Scopus | | Cited By in Scopus (108)

Santoni et al., 1996 G. Santoni, M. Perfumi, A.M. Bressan and M. Piccoli, Capsaicin-induced inhibition of mitogen and interleukin-2-stimulated T cell proliferation: its reversal by in vivo substance P administration. J. Neuroimmunol. 68  (1996), pp. 131–138.

Sathianathan et al., 2003 V. Sathianathan, A. Avelino, A. Charrua, P. Santha, K. Matesz, F. Cruz and I. Nagy, Insulin induces cobalt uptake in a subpopulation of rat cultured primary sensory neurons. Eur. J. Neurosci. 18  (2003), pp. 2477–2486.

Serreze et al., 1994 D.V. Serreze, M. Prochazka, P.C. Reifsnyder, M.M. Bridgett and E.H. Leiter, Use of recombinant congenic and congenic strains of NOD mice to identify a new insulin-dependent diabetes resistance gene. J. Exp. Med. 180  (1994), pp. 1553–1558.

Trudeau et al., 2000 J.D. Trudeau, J.P. Dutz, E. Arany, D.J. Hill, W.E. Fieldus and D.T. Finegood, Neonatal beta-cell apoptosis: a trigger for autoimmune diabetes?. Diabetes, 49  (2000), pp. 1–7.

Van Buren et al., 2005 J.J. Van Buren, S. Bhat, R. Rotello, M.E. Pauza and L.S. Premkumar, Sensitization and translocation of TRPV1 by insulin and IGF-I. Mol Pain, 1  (2005), p. 17.

van de Wall et al., 2005 E.H. van de Wall, D.X. Gram, J.H. Strubbe, A.J. Scheurink and J.M. Koolhaas, Ablation of capsaicin-sensitive afferent nerves affects insulin response during an intravenous glucose tolerance test. Life Sci. 77  (2005), pp. 1283–1292.

Verdaguer et al., 1996 J. Verdaguer, J.W. Yoon, B. Anderson, N. Averill, T. Utsugi, B.J. Park and P. Santamaria, Acceleration of spontaneous diabetes in TCR-beta-transgenic nonobese diabetic mice by beta-cell cytotoxic CD8+ T cells expressing identical endogenous TCR-alpha chains. J. Immunol. 157  (1996), pp. 4726–4735.

Verdaguer et al., 1997 J. Verdaguer, D. Schmidt, A. Amrani, B. Anderson, N. Averill and P. Santamaria, Spontaneous autoimmune diabetes in monoclonal T cell nonobese diabetic mice. J. Exp. Med. 186  (1997), pp. 1663–1676.

Wang et al., 2003 H. Wang, M. Yu, M. Ochani, C.A. Amella, M. Tanovic, S. Susarla, J.H. Li, H. Yang, L. Ulloa and Y. Al-Abed, et al. Nicotinic acetylcholine receptor alpha7 subunit is an essential regulator of inflammation. Nature 421  (2003), pp. 384–388.

Winer et al., 2001 S. Winer, I. Astsaturov, R. Cheung, L. Gunaratnam, V. Kubiak, M.A. Cortez, M. Moscarello, P.W. O’Connor, C. McKerlie, D.J. Becker and H.M. Dosch, Type I diabetes and multiple sclerosis patients target islet plus central nervous system autoantigens; nonimmunized nonobese diabetic mice can develop autoimmune encephalitis. J. Immunol. 166  (2001), pp. 2831–2841.

Winer et al., 2003 S. Winer, H. Tsui, A. Lau, A. Song, X. Li, R.K. Cheung, A. Sampson, F. Afifiyan, A. Elford and G. Jackowski, et al. Autoimmune islet destruction in spontaneous type 1 diabetes is not beta-cell exclusive. Nat. Med. 9  (2003), pp. 198–205.

Zhang et al., 2002 Y. Zhang, B. O’Brien, J. Trudeau, R. Tan, P. Santamaria and J.P. Dutz, In situ beta cell death promotes priming of diabetogenic CD8 T lymphocytes. J. Immunol. 168  (2002), pp. 1466–1472.

Supplemental Data

Corresponding author

7These authors contributed equally to the work.

Copyright © 2006 Elsevier Inc. All rights reserved.