Scanning electron microscopy of gastroscopic biopsies Ebrahim Fallah, MD Bernard M. Schuman, MD John H. L. Watson, PhD Jessica Goodwin, BS Detroit, Michigan
Endoscopic biopsy specimens of normal human gastric mucosa and mucosa in erosive and chronic gastritis were studied by scanning electron microscopy. The orifices of the gastric pits and the cobblestone surface of epithelial cells covered with villus-like and bulbous projections were observed. A possible mechanism for mucus secretion from these cells is suggested. The mucosa showed striking morphologic surface differences from the normal in cases of both erosive and chronic gastritis.
T
he gastric mucosa is unique among living surfaces for it must cope with and maintain its integrity under a wide variety of environmental conditions. It is affected by temperature, osmolality, mechanical trauma and differing ingested substances, including both food and drugs. Its inherent ability to withstand the action of hydrochloric acid and proteolytic enzymes still remains a mystery. In addition, the exact mechanisms responsible for gastric mucosal injury in various pathologic states are not yet fully understood 102 , although it is known that the gastric epithelium plays a major role in the formation of the so-called mucosal barrier. 3 We have employed the scanning electron microscope in studies of gastroscopic biopsies as an aid toward the elucidation of the nature and structure of this unique mucosal surface and in an attempt to understand more fully its alterations to injury.
MATERIALS AND METHODS Forceps biopsies of gastric mucosa were obtained from 4 normal human stomachs, from 3 patients with erosive gastritis, and from 3 with chronic non-specific gastritis. The cases were selected from among patients referred to the gastrointestinal endoscopy unit for evaluation of a wide variety of conditions. Endoscopic diagnosis was used for the initial classification which was later confirmed by light microscopy'. Three biopsies with an average diameter of 1 mm were taken in each case: 2 from the body and 1 from the antrum of each stomach. One body biopsy was paraffin-sectioned and H&E stained for conventional light microscopy. The other and the antrum biopsy were put immediately into cacodylate-buffered, 2%
glutaraldehyde solution for fixation. One half of each of these was critical-point dried in C02 and subsequently vacuumcoated with gold-palladium for scanning electron microscopy. The other half of each was processed for transmission electron microscopy by conventional methods and without critical point drying, using Araldite for embedment. In addition, several of the critical-point dried specimens were reprocessed for transmission after examination by scanning electron microscopy. It has long been known that shortening the exposure time to outside environment by immediate fixation of the tissue as it leaves the donor reduces autolytic processes to a minimum, and preserves the surface as closely as possible in its in vivo condition. Concurrent experiments to determine the effects of washing and other preparative techniques upon canine stomach mucosaS (e.g., rinsing in normal saline, ice water, or 1% HCI), showed that any treatment or manipulation before fixation is capable of introducing surface alterations. Consequently, every sample examined in this study was immediately immersed in the fixative without pretreatment of any kind. It was found, when the specimen surface was covered by an original thick layer of mucus after fixation, the mucus could be carefully teased from the surface mechanically, leaving the cell surface undamaged. The antrum biopsies were examined for each case by scanning and for most by transmission electron microscopy without noting any surface differences relative to the specimens taken from the body of the same stomach. Consequently all of the following report concerns the body of the stomach.
From the Division of Gastroenterology, Henry Ford Hospital, and the Department of Physics and Biophysics, Edsel B. Ford Institute for Medical Research, Detroit, Michigan. Reprint requests: John H. L. Watson, PhD, Department of Physics and Biophysics, Edsel B. Ford Institute for Medical Research, 2799 West Grand Blvd., Detroit, Michigan 48202. VOLUME 22, NO.3, 1976
137
138
GASTROINTESTINAL ENDOSCOPY
Although it is always difficult and usually speculative to draw dynamic conclusions from the 2-dimensional images of static electron micrographs, what could be all stages in the formation of these projections by an extrusion process from the cell cytoplasm can be visualized in Figures 3 and 4, from a first nippl ing of the epithelial surface, through an intermediate stage of larger maturing "particles", to the final end-product of spherical "particles" raised well above the surface. While some of the properties of these bulbous structures are similar to those of microvilli (they are of similar diameter and by transmission electron microscopy of thin sections sometimes have a root-like structure) (as in Figure 14A), neither their peculiar shape, short length, nor the apparent process of their formation are usual characteristics of microvilli. Many of the projections are shaped like thumbs (clear arrow, Figure 4,) abou~ 1,5% wider at their rounded extremity than at their base but without a narrow supporting stalk. As such they resemble true microvilli more closely, but are still connected laterally in
Endoscopic biopsy specimens of normal human gastric mucosa and mucosa in erosive and chronic gastritis were studied by scanning electron microscopy. The orifices of the gastric pits and the cobblestone surface of epithelial cells covered with villus-like and bulbous projections were observed. A possible mechanism for mucus secretion from these cells is suggested. The mucosa showed striking morphologic surface differences from the normal in cases of both erosive and chronic gastritis.
T
he gastric mucosa is unique among living surfaces for it must cope with and maintain its integrity under a wide variety of environmental conditions. It is affected by temperature, osmolality, mechanical trauma and differing ingested substances, including both food and drugs. Its inherent ability to withstand the action of hydrochloric acid and proteolytic enzymes still remains a mystery. In addition, the exact mechanisms responsible for gastric mucosal injury in various pathologic states are not yet fully understood 102 , although it is known that the gastric epithelium plays a major role in the formation of the so-called mucosal barrier. 3 We have employed the scanning electron microscope in studies of gastroscopic biopsies as an aid toward the elucidation of the nature and structure of this unique mucosal surface and in an attempt to understand more fully its alterations to injury.
MATERIALS AND METHODS Forceps biopsies of gastric mucosa were obtained from 4 normal human stomachs, from 3 patients with erosive gastritis, and from 3 with chronic non-specific gastritis. The cases were selected from among patients referred to the gastrointestinal endoscopy unit for evaluation of a wide variety of conditions. Endoscopic diagnosis was used for the initial classification which was later confirmed by light microscopy'. Three biopsies with an average diameter of 1 mm were taken in each case: 2 from the body and 1 from the antrum of each stomach. One body biopsy was paraffin-sectioned and H&E stained for conventional light microscopy. The other and the antrum biopsy were put immediately into cacodylate-buffered, 2%
glutaraldehyde solution for fixation. One half of each of these was critical-point dried in C02 and subsequently vacuumcoated with gold-palladium for scanning electron microscopy. The other half of each was processed for transmission electron microscopy by conventional methods and without critical point drying, using Araldite for embedment. In addition, several of the critical-point dried specimens were reprocessed for transmission after examination by scanning electron microscopy. It has long been known that shortening the exposure time to outside environment by immediate fixation of the tissue as it leaves the donor reduces autolytic processes to a minimum, and preserves the surface as closely as possible in its in vivo condition. Concurrent experiments to determine the effects of washing and other preparative techniques upon canine stomach mucosaS (e.g., rinsing in normal saline, ice water, or 1% HCI), showed that any treatment or manipulation before fixation is capable of introducing surface alterations. Consequently, every sample examined in this study was immediately immersed in the fixative without pretreatment of any kind. It was found, when the specimen surface was covered by an original thick layer of mucus after fixation, the mucus could be carefully teased from the surface mechanically, leaving the cell surface undamaged. The antrum biopsies were examined for each case by scanning and for most by transmission electron microscopy without noting any surface differences relative to the specimens taken from the body of the same stomach. Consequently all of the following report concerns the body of the stomach.
From the Division of Gastroenterology, Henry Ford Hospital, and the Department of Physics and Biophysics, Edsel B. Ford Institute for Medical Research, Detroit, Michigan. Reprint requests: John H. L. Watson, PhD, Department of Physics and Biophysics, Edsel B. Ford Institute for Medical Research, 2799 West Grand Blvd., Detroit, Michigan 48202. VOLUME 22, NO.3, 1976
137
138
GASTROINTESTINAL ENDOSCOPY
Although it is always difficult and usually speculative to draw dynamic conclusions from the 2-dimensional images of static electron micrographs, what could be all stages in the formation of these projections by an extrusion process from the cell cytoplasm can be visualized in Figures 3 and 4, from a first nippl ing of the epithelial surface, through an intermediate stage of larger maturing "particles", to the final end-product of spherical "particles" raised well above the surface. While some of the properties of these bulbous structures are similar to those of microvilli (they are of similar diameter and by transmission electron microscopy of thin sections sometimes have a root-like structure) (as in Figure 14A), neither their peculiar shape, short length, nor the apparent process of their formation are usual characteristics of microvilli. Many of the projections are shaped like thumbs (clear arrow, Figure 4,) abou~ 1,5% wider at their rounded extremity than at their base but without a narrow supporting stalk. As such they resemble true microvilli more closely, but are still connected laterally in
