The TRP superfamily of ion channels is a structurally defined group of cation channels showing an enormous functional diversity [1
]. Although several studies have shown the distribution and tissue specific expression of these channels in mammals, the relative importance of individual channels for certain tissues or cell types is widely unknown in most cases. The present study systematically quantifies the expression of all members of the TRPC, TRPV, TRPM and TRPA subfamilies in a broad spectrum of tissues.
Our results suggest a high relevance of TRPM channels in many murine tissues, since members of this subfamily showed the highest expression levels. These findings may qualify some studies that focused on members of the TRPC subfamily and tried to correlate gene expression with physiological function. Among the TRPMs, the chanzyme TRPM7 dominated over all other tested TRP members and seems to be ubiquitously expressed. As TRPM7 has a unique function for cellular Mg2+
] its wide distribution in all tested tissues looks plausible. In most brain areas the copy number of TRPM3 mRNA exceeded even that of TRPM7. In contrast to humans where TRPM3 shows a high expression in kidney [17
], mouse kidney did not reveal extensive expression of TRPM3. This is in agreement with the results of Grimm and coworkers [18
], who studied the tissue distribution of TRPM3 using northern blot analysis and immunochemical methods. Obviously, the osmotically regulated channel TRPM3 has different physiological relevance in different species.
Members of the TRPC subfamily were also widely distributed in murine tissues, but their overall expression was about one order of magnitude lower than that of the TRPMs. The dominant and most widely distributed isoform is TRPC3, which occurred in high concentrations in all tested brain areas. In addition, TRPC3 expression was obviously present in many other tissues such as liver, lung, kidney, ovary and testis. This is largely in agreement with results from rat tissues [11
], only the absence of TRPC3 from rat liver showed a marked difference to our results. Further, we did not observe the strong expression of TRPC1 shown by Garcia and Schilling [11
]. In many parts of human brain TRPC3 showed an at least tenfold higher expression than in peripheral tissues [12
], a finding that is supported by this study in the mouse. Riccio and coworkers [12
] also showed a high prevalence of TRPC1, C4, C5, and C7 in the CNS, i.e. a much stronger expression in all brain regions than in peripheral tissues. Only TRPC6 expression was in the same range in CNS as in peripheral tissues. Our study comes to similar results for the mouse. Furthermore, as levels of TRP channel mRNAs are normalized to that of 18S rRNA (Fig. ), it is possible to estimate the relative importance of an individual TRP member for a certain tissue or brain region. For the cellular localization of TRP mRNAs in the CNS we used the in situ
hybridization technique. In agreement with two previous publications [19
] we detected TRPC3 mRNA and TRPC6 mRNA preferentially in Purkinje cells, neurons of basal ganglia and ventricle ependym, while TRPC5 mRNA was only found in a very small amount in any of these cell types. The morphometric results of the mRNA levels for TRPC3, C5 and TRPC6 displayed a good correlation between Real-time RT-PCR and in situ
Members of the TRPV subfamily showed an overall lower expression compared to members of TRPC and TRPM subfamilies. This is not surprising, since several of these channels serve as molecular sensors for pain and heat sensation and occur preferentially in sensory nerve terminals [1
]. However, some characteristic exceptions have to be mentioned. High expressions of TRPV4 and TRPV5 were observed in kidney (Fig. ). This finding is in agreement with earlier studies showing the importance of TRPV5 for calcium re-absorption in the kidney [21
TRPA1 expression was even lower than that of TRPVs and more or less restricted to nervous tissue, ovary, testis and spleen. Low expression of TRPA1 in brain is compatible with its presumed function as a neuronal sensor molecule. Its occurrence in reproductive organs is described for the first time in this study.
Tissues or organs that mainly consist of muscle cells such as the aorta, skeletal muscle and heart showed an overall low expression of TRP channels. However, calcium conducting TRP channels may also be of great importance for muscle function, since contraction is controlled by calcium ions. The modulation of muscular calcium signaling by increased calcium entry through the sarcolemma can have consequences for muscle function and may cause dysfunction and disease [9
]. For members of the TRPC subfamily the expression in smooth muscle and tissue preparations such as the aorta have been well investigated. Signals of TRPC1, C3, C4, C5 and C6 have been reported in mouse and rat aorta, while TRPC2 and C7 transcripts were not observed [7
]. Our study confirms many of these results. In addition, we observed the presence of TRPC2 in the aorta, but very low signals for TRPC4 and C5. Furthermore, other members of the TRP family may be of relevance for aortic smooth muscle. We found TRPV4 expression almost as high as that of the most abundant TRPC member, TRPC3. TRPC3 and TRPV4 expression was even exceeded by TRPM5 and TRPM7 (Fig. ).
In human heart muscle several TRPC transcripts have been detected. TRPC1, C4, C5, and C6 mRNAs show moderate levels compared to CNS [12
]. This study comes to the result that among TRPC family members only TRPC1, C3 and C6 occur in mouse heart at significant levels. We found only traces of TRPC4 and C5 mRNAs. Differences between mouse and human heart or the fact that TRPC4 and C5 are not constitutively expressed may be responsible for these findings. Recently, the upregulation of TRPC4 and C5 was shown in cardiomyocytes in response to inhibition of SERCA expression. In this model TRPC4 and C5 seem to be induced after insufficient function of SR calcium accumulation [22
]. Using in situ
hybridization experiments we were able to localize the mRNAs for TRPC3 and TRPC6 in cardiac muscle cells but not in the endothelium of cardiac blood vessels. It is known that the subgroup of TRPC3/C6/C7 can assemble into homo- and heterotetramers [23
]. As TRPC3 and C6 are expressed at significant levels in mouse heart, these two channels have the potential to form several Ca2+
conducting non selective cation channels in myocytes and contribute to cardiac Ca2+
Surprisingly, cardiac expression levels of TRP mRNAs showed considerable variability among different mouse strains. In C57Bl mice TRPC3 was dominantly expressed in heart muscle whereas TRPC6 was the major isoform in NOD and Balb/c hearts. Heterogeneity was also observed for TRPV2 expression. This mRNA was only found in NOD and Balb/c hearts and almost absent in the C57Bl strain. Recently, the presence of TRPV2 in the sarcolemma of cardiac muscle has been reported. Further, the cardiac-specific overexpression of TRPV2 caused a cardiomyopathy due to cellular Ca2+
overload in a transgenic mouse model. The severity of the cardiomyopathy was roughly related to sarcolemmal TRPV2 levels [24
]. These data show that TRPV2 can be an important element of cardiac Ca2+
homeostasis. Besides TRPCs and TRPV2, other members of the TRP family were found to be expressed in heart muscle, e.g. TRPV4, V6, M4 and M7. This is in accordance with previous studies showing the expression of TRPV4 [25
], TRPV6 [26
]and TRPM4 [4
] in murine heart.
In addition to earlier studies that focussed on the TRPC subfamily [27
], we investigated expression levels of TRPV and TRPM channels in skeletal muscle. Two reports described the presence of TRPC1, C2, C3, C4 and C6 mRNAs by standard RT-PCR and showed that the corresponding channel proteins are localized in the sarcolemma [27
]. These data are in agreement with our results, however we suggest a dominant role of TRPC3 in mouse muscle, since the TRPC3 transcript occurs in much higher concentrations than those of the other TRPC members. In addition, mRNAs of TRPV3, V4 and V6 as well as TRPM3, M4 and M7 were found in considerable concentrations in several mouse skeletal muscles including diaphragm.
Apart from RT-PCR studies, databases containing expressed sequence tags (ESTs) can be used to estimate gene expression. The relation of the ESTs of a certain gene to the total number of ESTs derived from a tissue provides a measure for gene expression . However, TRP transcripts code for membrane proteins of low abundance and constitute only a small fraction of the cellular transcriptome. Therefore it is not surprising that ESTs derived from TRP RNAs occur at very low frequencies. EST databases are not very informative for tissues such as skeletal muscle and heart, since the number of hits for individual TRP transcripts is often 0 or 1. Large numbers of ESTs exist for certain brain regions. In combination with the fact of a rather high TRP channel expression in nervous tissue EST frequencies may reflect gene expression and can be compared to RT-PCR data. For cerebrum and cerebellum our RT-PCR data (Fig. ) widely agree qualitatively with the occurrence of EST frequencies included in UniGene. Both sets of data show high expression most TRPC members, TRPM3, 4 and 7, followed by TRPV2, 4 and 6. However, current EST data show quantitative differences to the RT-PCR data and reveal some unexpected negative results. For example, in UniGene there is no hit for TRPC1 in cerebrum and cerebellum is negative for TRPV4 and 6. Thus, to date RT-PCR seems to be the superior method for studying gene expression of low abundance transcripts such as TRP channel mRNAs.