Corticotropin-releasing factor immunostaining in the rat spinal cord and medulla oblongata: An unexpected form of cross-reactivity with substance P

Abstract

By use of two antisera (α-CRF A, α-CRF B) raised against conjugates of o-CRF bovine thyroglobulin, cryostat sections of formaldehyde-fixed gelatin models containing o-CRF can be stained. The staining intensity was quantitated by use of an automated microfluorimeter and was shown to be dependent on the concentration of o-CRF (1–300 μM) added to the gel. Determination of the CRF staining intensity after incorporation of o-CRF-related peptides and fragments indicated that both antisera reacted with the C-terminal region of o-CRF. They showed poor cross reactivity with r-CRF fixed in the gel. In the same models, r-CRF could be immunostained efficiently by use of an antiserum (α-CRF C) raised to a conjugate of r-CRF and thyroglobulin. This antiserum reacted with the N-terminal and midportion parts but not with the C-terminal fragment of o-CRF fixed in the gels. By use of both o-CRF antisera nerve fibers can be stained in the rat hypothalamus (median eminence) and in the medulla oblongata (spinal trigeminal tract and nucleus) and spinal cord (dorsal horn). Immunoinhibition experiments showed that o-CRF caused a concentration-dependent quenching (0.001–1 μM) of the immunostaining of o-CRF-containing models, rat median eminence and medulla oblongata preparations. α-CRF C also stained CRF immunoreactive (CRFi) fibers in the rat hypothalamus with an equal distribution to that found with the o-CRF antisera. However, no immunostaining was found in the spinal trigeminal nucleus and tract and in the dorsal horn, indicating that these fibers store different CRF-related products from those found in the hypothalamus. The CRFi in the medulla oblongata and spinal cord induced by α-CRF A was completely abolished 1 week after treatment of adult rats with capsaicin, a substance known to deplete Substance P (SP) from those areas. Gels incorporated with SP showed a concentration-dependent increase (range 10–1000 μM) in immunostaining with both o-CRF sera but not with the r-CRF antiserum. In addition, incubation of o-CRF sera with SP caused a concentration-dependent quenching (range 10–100 μM) of immunostaining in SP-containing models. SP at a concentration of 100 μM was also effective in quenching the CRFi in the dorsal horn and spinal trigeminal area. Quenching was also obtained with the C-terminal part of o-CRF (range 0.002–0.1 μM), which indicates that both CRF antisera contain an immunoglobulin which recognizes determinants on CRF as well as on SP. We conclude that CRFi in primary sensory nerves in the medulla oblongata and spinal cord may be due to cross reaction of CRF antisera with SP, located in high concentrations in such areas. In addition, the data show that the non-biological gelatin model is a good tool to investigate unexpected cross reactions in immunocytochemistry.