Commentary on “Classifications of Anterior Pituitary Cell Types With Immunoenzyme Histochemistry”: Dr. Paul Nakane Blazed the Trail to Modern Technology

Commentary

If the Histochemical Society had produced a time capsule in 1970, the article by Dr. Paul Nakane¹ would no doubt have been included. It contains technical and conceptual breakthroughs that became turning points for many researchers in the field. Dr. Nakane had reported the development of technology earlier, which used conjugates of peroxidase and antibody to detect the site of antigens by the peroxidase reaction product.^2,3 The 1970 article reported the application of this technology to the detection of pituitary cells at light and electron microscopic levels,¹ showing significant feasibility. My commentary will focus on the immediate impact of the article and its lasting impact, which may explain why it is cited today.

The immunoperoxidase approach initially had multiple advantages over the currently available immunofluorescent protocols: (1) one did not need a fluorescent microscope as the peroxidase substrate could be detected under bright-field microscopy; (2) reaction products could also be detected at the electron microscopic level; and (3) the reaction products from two different substrates, distinguished by color, allowed for the sequential detection of two different antigens in the same cells, section, or tissue. This technology was ideal for those studying specific anterior pituitary cell populations, which contain a mixture of cell types. For four decades, pituitary cytophysiologists’ approaches to cell differentiation were limited to tinctorial stains, or the size and shape of secretory granules, neither of which identified cells by the hormone produced.⁴

Imagine a scientific world without access to immunoperoxidase technology! This was the world I faced as a graduate student challenged by my advisor Dr. Nicholas S Halmi to identify and study the rodent anterior pituitary corticotrope.⁵ I was not alone. During the same period, Dr. Sam Spicer gave Dr. Robert Phifer the same challenge in studies of human pituitaries,⁶ and Dr. Edward G. Rennels challenged Dr. Phyllis Bowie to do the same in rodents.⁷ Drs. Halmi, Spicer, and Rennels had just come from a national meeting where the identity of the elusive corticotrope in rodents was being hotly debated. We three students turned to this novel immunoperoxidase technology, and all of us were successful.

The 1970 article by Nakane¹ was for us a beacon, shining a light on what was feasible. We were guided by the detailed methodologies and could make choices accordingly. Greeted by photographs of immunolabeling applied to each of the anterior pituitary cell types, we could apply the technology to thick sections or ultrathin sections, the latter of which appeared to have the best morphology. In fact, the electron micrographs (EMs) of immunolabeled somatotropes and lactotropes are state of the art even by today’s standards. The EM photographs of labeled thick sections of thyrotropes and corticotropes did not resolve the questions about the size or distribution of secretory granules, which was at the heart of the controversy about their identity.⁸ For these reasons, I elected to focus on the use of ultrathin sections based on Nakane’s protocol. I recall that one of his techniques turned out to be a breakthrough as it alleviated the notorious stickiness of the peroxidase substrate to plastic. This could be mitigated by immersing the EM grids in a swirling peroxidase substrate solution.¹

Nakane also reported findings at both the light and electron microscopic levels that are relevant to today’s studies of pituitary networking and cytophysiology. The application of his dual peroxidase labeling showed that gonadotropes could also be bihormonal, containing both gonadotropins, luteinizing hormone (LH) and follicle-stimulating hormone (FSH), which also challenged the original one-cell-one hormone dogma in the field. This is now the recognized state of most gonadotropes, and mechanisms underlying the non-parallel release of LH and FSH from this cell type are still being investigated.⁹

In addition, the paper reported that gonadotropes have diverse morphologies. Nakane illustrated a subset of FSH cells that were stellate with peripheral granules (type B FSH cells). As it turns out, we showed that this form may represent multihormonal cells producing both FSH or LH and adrenocorticotropin.¹⁰ The significance of multihormonal pituitary cells is now under active investigation,¹¹ most recently studied with single-cell RNA-sequence approaches.¹² Many researchers believe that these cells support pituitary transdifferentiation and plasticity.

The dual labeling allowed detection of the differential distribution of the pituitary cell types and geographical associations between cell types, such as somatotropes and corticotropes. There was a close physical association between lactotropes and gonadotropes, which may predict important paracrine regulatory influences. The significance of these associations predicts important networking relationships, which are still being investigated.^13,14 For example, lactotropes may be dependent on gonadotropes for their development through paracrine interactions or networking.¹⁵

The concluding paragraphs discuss the significance of pituitary cell heterogeneity in storage, suggesting that they are in different physiological states. This important observation has been confirmed many times. Some actively secreting cells may lose their hormone stores and are identified only by their content of mRNA. Also, some cells may be stem or progenitor cells, not yet storing sufficient hormone for identification by immunolabeling. These considerations are conceptually important as we interpret results today.

Finally, whereas there is no question that the paper was revolutionary for pituitary cytophysiologists, the immunoperoxidase technology exploded in the 1970s as researchers studying all types of tissues sought to identify their favorite cell type. Today, we have many choices of immunolabeling tools, including a return to sensitive immunofluorescence methods. However, immunoperoxidase with its different colorful substrates and highly sensitive output remains widely used.