A cabinet, an egg and an antibody

HERCULES and Hippolyta are the J. Paul Getty Museum’s most famous heavy lifters. Carved in oak and swathed in gold, they are also impressive double-taskers. As Greek mythological characters, the sculptures personify strength and bravery. As the base of a towering French cabinet, they dig in their heels, flex their muscles and hoist a chest of drawers decorated with scenes of Louis XIV’s military victories.

It’s a big job, and the pair has been doing it heroically since about 1680, when the cabinet emerged from the workshops of furniture maker Andre-Charles Boulle and gilder Jean Varin. Standing 7 1/2 feet tall and veneered with pewter, brass, tortoiseshell, ebony, ivory and wood marquetry, the cabinet is an over-the-top creation that compels museum visitors to stop and stare.

But -- art being art -- every last detail has to look exactly right. And that’s a problem for Hercules and Hippolyta. Scholars have long admired the cabinet’s workmanship but debated the sculptures’ skin color. Should it be so startlingly white? Or should it be closer to the imitation bronze tones on a nearly identical cabinet at Drumlanrig Castle in Scotland?

After a two-year investigation -- with broad implications for the field of art conservation -- the puzzle has been solved by an unlikely couple: Arlen Heginbotham, associate conservator of decorative arts and sculpture at the Getty Museum, and Michael Quick, dean of research and professor of biological sciences at USC. And therein lies a tale of art and science, proteins and antibodies, enzymes and immuno-fluorescence microscopy.

When the Getty purchased the cabinet in 1977, the skin of the figures was a nearly black simulation of bronze. During a thorough restoration in London, before the piece was shipped to Los Angeles, it was determined that the dark tone on the sculptures was the result of several layers of paint applied many years after the cabinet was made. Taking the examination further, the restorer excavated the underlying layers, concluded that the original color of the figures was white, in imitation of marble, and stripped off all but that first layer of material.

But some specialists thought the white looked wrong, and tests have long since determined that it is a chalk-based gesso meant to be a primer, not the finished surface.

The error hasn’t prevented the cabinet from becoming a widely published keystone of the Getty’s decorative arts collection. It has been featured in the museum’s guidebooks for many years. Other works by Boulle have brought as much as $5.7 million at auction.

“This is one of our great masterpieces,” Heginbotham said. “Because it’s so important, we wanted to understand the materials that had been used and the working technique as well as we could.”

That wouldn’t be easy. Records of the 1977 analysis are extremely sketchy. The best chance of gaining insight into the matter was to study the figures on the Scottish cabinet. They are thought to have been finished like those at the Getty. With permission from the Scottish authorities, Heginbotham cut tiny cross sections of paint from the sculptures, along cracks where extractions would cause no harm. Back in Los Angeles, he analyzed them, with the help of scientists at the Getty Conservation Institute.

Using standard methods for identifying art materials, such as gas chromatography, mass spectrometry and electron microscopy, they found that the samples contained six original layers: gesso, iron-based red pigment, vermilion pigment, a transparent organic material, a mixture of lead white and vermilion pigments, and copper-based pigment. They also determined that three additional layers had been applied later.

“Our traditional conservation techniques gave us quite a bit of information,” Heginbotham said. “What they couldn’t tell us was what this organic layer in the middle was. That was the mystery. We didn’t know why it was there, what function it was serving. If we wanted to understand what the sculptures really looked like, we had to be able to do real replications and understand the working technique.

“We tried several ways to figure it out, and we couldn’t do it. The biggest problem is that it’s buried between two other layers and it’s so thin. It’s 5 microns thick. We couldn’t isolate it. We couldn’t get a pure sample to analyze it. We were stumped.”

He made a series of tests, trying different materials sandwiched between the known quantities and comparing microscopic cross sections of the samples with those from the Scottish cabinet. But he couldn’t emulate the mystery component -- an extremely thin, transparent layer that hugged contours of the underlying pigment. Varnishes tended to flow out into a smooth surface; glues soaked in and didn’t form a coherent layer.

Eventually he hit upon the possibility of egg white, sometimes used as a sealer. That idea produced a “eureka” moment. The physical appearance was right. But Heginbotham had to prove that egg white actually was the mystery material.

While camping, a clue

WITH the problem roiling around in his head, he went on vacation with his wife, conservator Leslie Rainer, and her sister, biologist Julia Rainer.

“I was explaining this to my sister-in-law because I talk about dorky things, even when I am on vacation,” he said. “She said, ‘Well, if it’s egg white, it’s a protein. You should be able to identify it very easily with antibodies.’

“Antibodies are something we have no exposure to in art conservation science,” Heginbotham said. “So she put me in touch with her old professor, who put me in touch with somebody else at USC, who said, ‘Why don’t you call this bright new kid on the block, Michael Quick?’ ”

“So I got this nutty call,” said Quick, who claims he got a C-minus in college art appreciation and knows nothing about art conservation. “Arlen left this voice mail. I couldn’t even figure out what it was about because it was so out of the blue. It sounded crazy -- but I’m always up for these kinds of things because you never know where they are going to lead.

“Once Arlen had the idea that this layer was an animal product, there were lots of things we could do biologically,” he said. “Antibodies as a technique is basically limited to finding protein-based materials, but that means a lot because there are a lot of them, and it’s a highly specific approach. Another advantage is that you can detect tiny amounts of protein.

“The process is based on when your body gets an infection,” Quick said. “You get a cold; your body mounts an immunological response to fight this off. A virus comes in; it generates antibodies against proteins that reside on that virus. Antibodies are the internal defense system to get rid of foreign invaders, and they work by attacking proteins.

“The reason this works so well is that the antibody attacks just the protein associated with the virus. And you can use that technique to do something experimental. That is, if you have a protein of interest, you can create an antibody against it and have the antibody find it. That’s what we do medically all the time in experimental biology.”

Even so, Quick didn’t have high hopes for applying the technique to art conservation.

“The big concern I had was that we are talking about stuff made hundreds of years ago,” he said. “There might have been albumin proteins in the original material, but who knows whether there’s a chemical structure there now that an antibody recognizes? After 300 years, did the parts degrade away? There was no way this was going to work. I was so sure of it.”

Nonetheless, they decided to do a control study. To try out the antibody method on something old -- though not so aged as the cabinet -- Heginbotham and Quick clipped a tiny sample from a photograph on albumin-coated paper made about 150 years ago.

“We took a little snip, dissolved it and did our ELISA assay,” Quick said, using the acronym for Enzyme Linked Immunosorbent Assay. “Essentially, it’s the same procedure as in an AIDS test. In this case, you put a chip from a piece of artwork in a test tube and dissolve it in a water-based fluid that has a detergent in it. When you put some of that stuff in the well of a little dish, the protein sticks to the walls. Then you add an antibody that attacks the protein and a second antibody that recognizes the primary antibody and has a signal attached to it, so you can see it.” The signal, an enzyme, causes a chemical reaction that makes the clear fluid turn yellow.

They repeated the process with albumin from a hen’s egg and chips from Heginbotham’s tests. They also did negative control tests with other materials containing no protein and, finally, with the original paint and the later additions from the Scottish cabinet. The results can be seen on the walls of 1/2 -inch-wide wells in a clear plastic tray about the size of a postcard.

“We were able to get a positive response to albumin using an ELISA on the original chip,” Quick said.

“On a sample you could barely see, about 400 years old,” Heginbotham added.

“So,” Quick continued, “very cool, very exciting, a lot of fun. But there are problems. The biggest is that you have to dissolve the sample away. And you know protein is in your sample, but where in your sample? You don’t get any spatial information.”

“From our perspective,” Heginbotham said, speaking as a conservator, “we were already doing better than we could have done with our traditional techniques. But then comes the idea, what if we could push this further and say not only it is there but where it is?”

Quick had an answer for that too.

“In this ELISA case, you have a system where your secondary antibody is an enzyme,” he said. “Another way to do it is to attach to the secondary antibody a fluorescent molecule that glows. Instead of an enzyme signal, you look for a fluorescent signal that lights up.”

Using that procedure, immuno-fluorescence microscopy, or IFM, a cross section of paint layers from the Scottish cabinet was embedded in a disk of plastic resin, about the size of a quarter, with an edge of the sample exposed. Then beads of the primary and secondary antibody solutions were applied to the exposed area. When the disk was viewed through a microscope with ultraviolet light, the protein glowed bright red. The mystery layer in the paint chip was no longer a mystery.

“The beauty of this is that it takes so little equipment. The techniques that didn’t work at the Getty Conservation Institute take a $500,000 machine that would fill this whole room,” Heginbotham said of the laboratory at USC.

‘The right people, the right time’

OTHER conservators have tried to use antibodies for identifying art materials since 1962, he said, but they became discouraged. There weren’t enough antibodies available to do the work, but now there are.

The success of the project boils down to “the right people, the right time and the right problem,” said Marco Leona, David H. Koch scientist-in-charge in the department of scientific research at the Metropolitan Museum of Art in New York. “Conservators have been interested in antibodies and immunological techniques, but they haven’t been reliable. Arlen and Michael looked at it once again at exactly the time when the technique was ready in the biochem labs. It’s remarkable that there was a great cooperation between a conservator at the Getty and a biochem professor who was interested in applications that are not usual in biochemistry research. That’s important and something our field needs more of.”

Their work has many possible applications in the conservation of paintings, textiles, parchment and ethnographic objects as well as decorative arts, he said. “The beauty of the method they have developed is that it deals with very small samples. It lets you understand the protein and, better, where that protein came from. You can say something is cowhide or sheep hide or kangaroo hide by the distribution of the amino acids.”

At the Getty Conservation Institute, scientist Julie Arslanoglu is already coordinating a project that will begin to integrate the new method into more traditional procedures.

But what about Hercules and Hippolyta? Will their white skin be returned to its original reddish bronze?

“That will be a decision the curators have to make,” Heginbotham said, “and it will be a hard decision.”