Between the cells: control by glue.

The meshwork of protein and sugar molecules that holds together different layers of cells in the body also influences their structure, metabolism, behavior and development. To examine just how this extracellular matrix affects the cells attached to it, biologists are growing cells on laboratory plates, where the cell's semisolid support and surrounding solution can be manipulated directly. Now Lola Reid of Albert Einstein College of Medicine in New York City reports that by varying the semi-solid support, scientists can manipulate liver cells in tissue culture to mimic a liver's several physiological states. She and her colleagues are beginning to describe the mechanisms behind this control.

Coaxing liver cells to maintain their normal characteristics while growing in tissue culture was a challenge that Reid found "laborious, but straightforward," she said in a seminar last week at the National Institutes of Health in Bethesda, Md. She and her co-workers spent four years working out the mixtures of nutrients and hormones that would sustain these cells. In contrast, most tissue culture experiements employ cells derived from tumors, because normal cells generally lose their specialized characteristics or die in laboratory culture.

To better mimic a cell's natural environment, Reid began growing the liver cells, called hepatocytes, not directly on plastic plates, but on a gel of collagen, a class of fibrous proteins that make up the biological glue. Reid finds that the type of collagen put on the laboratory plate determines the cells' "differentiation profile."

Cells placed on type III collagen resemble a normal "quiescent" adult liver--the cells maintain their adult characteristics and do not reproduce. Cells placed on type IV collagen reproduce rapidly for several weeks, also maintaining their adult functions. These cells resemble a liver regenerating after a portion has been surgically removed. Finally, cells on type I collagen resemble cells in a "wounded" liver, for example the liver of a hepatitis patient or a liver repeatedly exposed to alcohol. These cells lose their adult functions, reverting to a fetal form. Reid also observed a surprisng "synergy." Cells growing on collagen required fewer hormones than those on plastic.

The same correlations between collagen type and cell characteristics have been determined by microscope examination of liver tissue. Tony Martinez of Hahnemann University in Philadelphia finds that hepatocytes are associated with each of these three types of collagen depending on the liver's condition. In a quiescent liver, the hepatocytes contribute type III collagen to the extracellular matrix. After part of the liver has been removed, the remaining hepatocytes produce type IV collagen. But in a wounded liver, the hepatocytes make predominantly type I collagen.

Other components of the extracellular matrix also influence the activity of liver cells, Reid reports. Among these components are long sugar chains, called glycosaminoglycans (GAGs). These are sometimes found attached to proteins, and then they are referred to as proteoglycans. GAGs or proteoglycans from liver extracellular matrix have dramatic effects on liver cells growing in tissue culture.

In the presence of either active GAGs or proteoglycans, the cells change shape and pack together tightly as in a normal liver. These matrix components also induce the cells to synthesize special membrane structures, called gap junctions, that allow electrical signals to pass from one cell to the next, a characteristic of normal liver cells.

The influence of the components of the extracellular matrix on regulation of liver cell growth and specialization involves the turning on of certain genes and the turning off of others. Some major problems in off of others. Some major problems in modern biology involve the underlying mechanisms of such developmental control. The liver cell studies demonstrate that in some cases this regulation relies on different rates on the first step in gene expression: the copying of a gene into mesenger RNA. But in other instances the control comes later, via differences in the rates at which messenger RNA molecules are broken down in the cell. Differences in messenger RNA degradation are often cited as a possible mechanism for control of gene expression, but there have been few examples demonstrated. For both types of control, Reid and her colleagues observe differences in the regulation of the "common" genes, active in all cells, and other genes that provide liver cells with their characteristic properties.