Our results demonstrate, for the first time, that primary neonatal CFs express transcripts of NGF/TrkA and NT-3/TrkC pairs, and of BDNF, a pattern similar to that of primary cardiomyocytes except for the very low/undetectable expression of NT-3 in the myocytes (). NGF had the highest expression of the three in both adult and neonatal cardiac cells, whereas BDNF expression declined and NT-3 expression increased from neonatal to adult cells.
NGF is the most studied NT in cardiac injury. Earlier findings showed that NGF, which signals via TrkA but not TrkB and TrkC 
, triggers prosurvival activity in primary cardiomyocytes through an autocrine mechanism 
. This activity may be beneficial in vivo
, as NGF gene therapy is cardioprotective in models of myocardial infarction 
and diabetes 
. However, those in vivo
studies were not designed not tell whether myocardial NGF secreted under physiological conditions and/or in response to stressors arise from cardiomycytes. Furthermore, although it has been established that cardiac injury, such as in hypoxemia/reperfusion or myocardial infarction, sharply augments cardiac NGF 
, nothing is known of the ligands and cell-surface recepor(s) that trigger NGF upregulation, except for endothelin-1, which regulates NGF expression in cardiomycytes 
. However, it remains unknown whether endothelin-1 alters cardiac NGF expression in fibroblasts, which express endothelin-1 receptors 
. Identification of the molecules that control NGF upregulation is clearly important to better understand cardiac function, response to injury, and design therapeutics to boost myocardial NGF, including by systemic administration.
invades the heart where it causes acute myocarditis that lasts a few months, followed by a chronic debilitating cardiomyopathy in ~30% infected individuals many years after the infection 
. Therefore, the possibility exists that T cruzi
/host interaction triggers reactions that reduce or prevent damage of infected tissues. This idea is backed up by our findings presented here, which demonstrate that T cruzi
triggers an exuberant production of cardioprotective NGF (more than 80-fold increase) in cardiac cells growing in culture, predominantly in fibroblasts (). This NGF boost response discriminates between T cruzi
recognition of TrkA and TrkC, as it requires the former but not the latter (), an in vitro
response analogous to that seen in chagasic hearts (). Reflecting the T cruzi
recognition of Trks via its PDNF, bacterially expressed PDNF (sPDNF) reproduces the effect of T cruzi
infection, for it selectively boosts NGF in fibroblasts outside () and inside their natural myocardial niche ().
Given the prosurvival action of NGF on cardiomyocytes and the vigorous upregulation of bioactive NGF selectively in fibroblasts in response to T cruzi
infection or sPDNF stimulation, it is reasonable to hypothesize that T cruzi
-cardiac fibroblast interaction protects cardiomyocytes via a NGF-dependent paracrine mechanism. This prediction proved to be correct as medium conditioned by T cruzi
infection or sPDNF stimulation of fibroblasts prevented cardiomyocyte death resulting from oxidative stress (hydrogen peroxide) (). It has long been known that oxidative stress kills cardiomyocytes 
, including in chagasic cardiomyopathy development 
Furthermore, NGF released following T cruzi
-fibroblast interaction may also protect cardiomyocytes against deleterious effects additional to the oxidative stress shown here, as judged by the findings of others demonstrating NGF protection of cardiomyocytes subjected to growth factor deprivation, hypoxia/reoxygenation, and angiotensin II stimulation 
. We therefore hypothesize that enhanced NGF secretion following T cruzi-
PDNF recognition of fibroblast-TrkA is a mechanism responsible for long term symtomless and relatively pathology-free T cruzi
homing in the heart.
It will be of importance to know the biological outcome of T cruzi binding to TrkA or TrkC on cardiomyocytes, and TrkC on fibroblasts. It will also be important to know whether endogenous TrkA-binding molecules such as NGF reproduce the T cruzi action in fibroblasts.
Finally, as PDNF binding to TrkA on fibroblasts releases NGF () which, in turn, protects cardiomyocytes against oxidative stress () and perhaps other insults 
, and, as intravenous administration of sPDNF boosts cardiac NGF (), our results offer a possible novel translational medicine opportunity for cardiomyopathies where NGF gene therapy has proven to be valuable in myocardial infarction 
and diabetes 
. PDNF administration might also be useful in preventing or slowing down cardiomyocyte degeneration. PDNF might be a better therapeutic than NGF because it 1) activates both prosurvival TrkA and TrkC while NGF activates only TrkA; 2) has a unique dual prosurvival activity by directly activating TrkA and TrkC outside cells and, in the intracellular milieu, pro-survival Akt kinase 
; 3) creates a positive TrkA signaling loop by robustly upregulating NGF; 4) has a relatively long half-life (detectable in the heart for >30 min after iv injection, see Fig. S2