Paul Keim: “We Were Surprised It Was the Ames Strain”


October 10, 2011
In October 2001, Northern Arizona University microbiologist Dr. Paul Keim identified that the anthrax used in the attack letters was the Ames strain, a development he described as “chilling” because that particular strain was developed in U.S. government laboratories. Keim’s lab would later serve as a repository for all the anthrax samples the FBI collected during its investigation. This is the edited transcript of an interview conducted on June 27, 2011.

[What were you thinking after 9/11? Were you worried about a bioterror attack?]

The idea that we would have a biological attack following 9/11 was not at all in my mind. What I was thinking was that the country was in crisis. In fact, we might have bigger and better things to worry about than just biological weapons.

The surprise was then in early October, we heard about this anthrax case in Florida. The first that I heard about it was an FBI scientist named Doug Beecher gave me a call and said, “There’s this gentleman in Florida who has a case of anthrax,” and he began to ask me questions. Doug and I had known each other for some time. I was one of the experts in the field of anthrax even before the letters went out. So Doug called me up, and we were talking about what could be the source of this anthrax that was affecting this guy. He walked me through about half a dozen different scenarios that they were exploring for how Robert Stevens had contracted anthrax, and we talked about the pros and cons of each of those.

And then he signed off, and he called me back the next day. I was communicating instantly with my colleagues at the Centers for Disease Control trying to find out more details and telling them what I knew about what was going on. And then it was on a Thursday, it was about noon, and Doug calls me up and says, “The sample’s in the air.” And I said, “What?” He said: “The sample’s in the air. It just lifted off in a small private jet, and it’s flying from Atlanta, Ga., to Flagstaff, Ariz.” …

“When we identified the Ames strain, [Robert] Stevens was still alive, and it was only a matter of a few minutes or an hour until he passed away. … Suddenly you have a sample in your hands that had killed a person.”

He was asking you to do what? What was your role going to be?

We had one of the largest collections of different types of anthrax, what we call strains of anthrax, from around the world. And this is what we call a reference laboratory, so we had reference material from around the world. We had developed a method of DNA fingerprinting that hadn’t been possible even just a couple of years before. We used methods that were brand-new in the field, and we applied them to anthrax. And we developed a genetic database, a DNA fingerprinting database for all these different types.

What the FBI wanted us to do was see if we could identify what was killing Robert Stevens, so the team got organized. We got all our agents; we got all of our equipment reserved. It’s a big laboratory, and we had to make sure that it had priority.

So then I went out and met the plane when it landed at the Flagstaff airport. … The ramp came down, and this woman came off the plane with a box. I went over to meet her, and she said, “Dr. Keim,” and I said, “Yes.” She said, “This is the anthrax.” And I said, “Oh.”

So we filled out a whole bunch of paperwork where I took custody of the sample. I put it in my car and I drove back to the laboratory, where we cracked it open, processed it, and then extracted the DNA from the sample. We worked all night. By the start of business the next day, we knew what type of anthrax it was. The sample that was killing Robert Stevens matched a strain called the Ames strain.

How surprised were you that it was Ames?

We were surprised it was the Ames strain. And it was chilling at the same time, because the Ames strain is a laboratory strain that had been developed by the U.S. Army as a vaccine-challenge strain. We knew that it was highly virulent. In fact, that’s why the Army used it, because it represented a more potent challenge to vaccines that were being developed by the U.S. Army. It wasn’t just some random type of anthrax that you find in nature; it was a laboratory strain, and that was very significant to us, because that was the first hint that this might really be a bioterrorism event. …

At that point, … what do you assume this might lead to?

At this point, certainly, a foreign source or Al Qaeda was still a very real possibility. The fact there’s a laboratory strain just meant that a person had done it. This strain was widely distributed around the United States. Sixteen or 17 different U.S. laboratories had this strain. We didn’t know this at that time. This was something that the FBI discovered later on. But the fact that it was in multiple laboratories, and even in international laboratories, we knew that. We knew that lots of laboratories had this strain.

So it could have been Al Qaeda. It could have been who knows who. But certainly it didn’t instantly point to the U.S. Army or to a U.S. scientist.

So what do you do next?  …

Doug Beecher had been very important in this whole process, so he was the first person I called. I had to tell him what the match was. And then we joined a conference call between the FBI and the Centers for Disease Control just a few minutes later. Then we began talking with the Centers for Disease Control, gave a report on what we’d found. They’d been doing exactly the same analysis. This was a method that I invented, but it was one that I transferred [to the CDC]. …

So we actually did that initial analysis in parallel between Atlanta and Flagstaff, and when we came back we had the same result. And we both said it’s Ames strain.

So this is the Stevens sample that you both tested. What happens next? When do you hear about the next attack? And how does this start playing out?

Well, understand that when we identified the Ames strain, Stevens was still alive, and it was only a matter of a few minutes or an hour until he passed away. … That’s very chilling. Suddenly you have a sample in your hands that had killed a person. I’m an academic; I’m not an FBI agent. This isn’t run-of-the-mill stuff for me. This was very much an unusual occurrence in my life. Yeah, it was all very chilling. …

Then when other things started to happen, the letters began going to other places, then suddenly it was ramped up. And it seemed like just as soon as we would get back to status quo, we’d calm down, something else would happen. Another sample would go out; another person would be infected. And then there would be sending them back to us. So we had this almost like the pony express sending these samples to us by these small corporate jets on a regular basis for about six weeks. As they sampled different post offices or different places, they would put the samples on a jet and fly them out to us.

We became the FBI’s analysis lab in those crisis days, because they didn’t have a lab that could do this. USAMRIID, the United States Army [Research Institute of Infectious Diseases] lab at Fort Detrick, could handle the samples for them, but they couldn’t do the DNA analysis. …

There were really two things going on in those early days in parallel. There was the public health effort, and the Centers for Disease Control was trying to understand the public health impact. The FBI was investigating the crime. And they were sharing examples. But the FBI wanted an independent laboratory that could go into court and testify that this was the result, so they set us up to do that.

Pretty quickly we had FBI scientists, FBI agents in some cases, in my laboratory, monitoring our people, monitoring what we were doing in such a way so that it was done in a way that we could go to court and testify.

Did the FBI scientists that you were dealing with, did they understand the science behind all this, or was this something that was developing as you all went forward?

It was both. Some of the scientists were the veterans of the human identification DNA fingerprinting efforts that had been led by the FBI, so some of these scientists were outstanding geneticists. And while they were new to the microbial world, they certainly understand DNA technology and were right there with us the whole time.

Other scientists were being recruited in from other activities. Anybody who had a Ph.D. I think got recruited in the investigation in those early days. Some of them were out investigating bank robberies and white-collar crime, and suddenly they were investigating the anthrax letters. Some of those individuals had to study real hard and get up to speed, but other cases there were some world-class scientists involved.

Tell me a little bit more about USAMRIID’s role in all this … and the back-and-forth between your lab and USAMRIID.

So the actual processing of the infectious material, the very dangerous powders, was done at USAMRIID. And we never saw those, thank goodness, because those are very dangerous things. We deal with live cultures, and those are much safer. The methods that we use to handle offer very little risk to the personnel, whereas the powders can disperse and get spread around. It’s something we just absolutely avoid in our laboratory situation.

So John Ezzell and other people at USAMRIID were handling these powders and culturing them. And as they did this, they would then make cultures or [a] small gross of these that would be in a test tube. Think of a test tube with a lid on it, and inside is some solid Jell-O. They would grow the bacteria on that. So there would be a tube with Jell-O in it, and they would grow the bacteria, the anthrax on that, and then they would mail that or put it on an airplane and send it to us. Later on in the investigation, when it wasn’t quite as time-sensitive, much of the transfer of evidence was, in fact, done by shipping through different shipping companies.

But it was John Ezzell who had handled the material and then would send it out to us. We would have constant phone calls going back and forth as well as a lot of paperwork to track this all. I didn’t have a cell phone before this, and I had to get a cell phone to basically be on call 24/7. And that level of attentiveness was necessary for about six months in those early days of the investigation. …

Bruce Ivins, who was he? How long had you known him? …

The anthrax research community is relatively small. When we would get together for an international meeting, we’d have about 150 people, and those were the people from all over the world. Maybe half of those people would be from the United States. So Bruce and I and John Ezzell, we were all part of this group. We knew each other very well.

I met Bruce first in about 1998 in England. He was at the international anthrax meetings, and so was I. He was talking about his vaccine work, and I was presenting my work on DNA fingerprinting of anthrax strains. Our work was for the first time identifying differences between different isolates of anthrax from around the world. Bruce was intrigued by this because he was very interested in whether his vaccine would protect animals against all different types of strains, and he had never had the capability for telling one strain from another.

The Ames strain was one type that came from North America, but we had hundreds and hundreds from Africa, from Asia, from Canada. Different parts of the world were represented in our collection. And Bruce saw this as a resource to help validate or invalidate his vaccine that he was doing. So we began collaborating after that meeting. We started exchanging strains. We would send him strains, or he would send us strains to be DNA-fingerprinted so we could categorize them.

We designed a study where we were going to do a panel of these different types of strains, and he was going to then use them in his animal challenge experiments. And of course the Ames strain was part of that. Bruce himself had sent us the Ames strain once or twice, and we received it from other laboratories as well. In 2001 we had seven different types of Ames strains in our laboratory, and some of those had come from Bruce himself.

His reputation?

Bruce was a well-known, well-respected member of the anthrax research community. He was doing this cutting-edge work on the development of new vaccines. He was trying to get us beyond the vaccine that had been in place for nearly 40 years.

The older vaccine had problems. You would end up with severe knots in your arms sometimes after getting a booster shot. I experienced this myself, and other people in my laboratory had gotten these very intense, almost like a golf ball underneath your skin. So that was the hope, was we’d have a more effective [vaccine], and a vaccine that didn’t have as many problems. Bruce was one of the leaders in that entire area. …

So early on, your role is to do the DNA analysis. … What are you doing? What is the goal? And how does that work out?

Our method of DNA fingerprinting was able to identify the Ames strain. But once we got to that point, we couldn’t take it any further. We couldn’t tell one type of Ames strain from another type, and that’s what we needed to do. There were maybe 15, 20 laboratories in the world that had the Ames strain. We had to tell which laboratory it came from, and we couldn’t do that.

So we took the analysis to the next level. We were working with Claire Fraser, who was at that time the president of TIGR [The Institute for Genomic Research in Rockville, Md.], and one of her young scientists named Tim Reid, and a second scientist named Jacques Revel. And with support from Rita Colwell, who was the head of the National Science Foundation at that time, we were able to get the money to go in and determine the entire sequence, every nucleotide in the genome of the anthrax that killed Robert Stevens.

And when we did that, and then we started comparing that to other types of anthrax, we weren’t seeing any differences from one laboratory to another. They were indistinguishable. And that isn’t what we needed to track this back to its original source. We needed to be able to tell one type of Ames strain from another. …

The genetic thing could have been at that point sort of a dead end.

That’s what I thought was, in fact we were stumped. …

Terry Abshire at USAMRIID comes across something. Explain what she finds, why it’s so important, and how that pushes the investigation forward.

So Terry Abshire and Pat Worsham had been working in this area for quite sometime. They were the world’s experts on morphological variation in Bacillus anthraces, and what they noticed is that when the spores from the letters were put out on these agar plates that we call petri dishes, they would grow up and form characteristic morphology, or a colony morphology, and that morphology then wasn’t constant. There were, in fact, some very rare variances inside of that. So there were some colonies that looked different. …

They would take those, and they would purify them, and they would look at them. And sure enough, that difference held true over and over and over again. And in the end they were able to track down multiple different morphological variances. And those morphological variances then seemed to be a clue that we could follow up.

Explain this in layman’s terms, very simple terms, that the spores themselves genetically were a match. … What is the morphology?

The standard procedure for sequencing the DNA of a bacterium was that you’d take the most common type of colony, and then you’d extract all its DNA, and then you would determine the sequence of that. And what you’re looking at there is really a mixture of all those cells.

We had actually been looking at the most common type, not these rare types. The rare types were basically one scraggly little tree out in the forest, and we didn’t see it. We saw the forest, and we didn’t see the different types of trees.

Terry Abshire and Pat Worsham did. They spotted those rare types out there, and they were able to then use those to come up with kind of a characteristic for the spores that were in the letters.

So the morphology was what?

The morphology is a different shape, slightly different color. It’s the shape and color of these colonies growing on the petri dishes that was the clue that eventually led back to the RMR-1029.

… What other kinds of things were you testing?

One of the most important things that was happening during this period of time was the FBI was pursuing every laboratory that had the Ames strain. They were putting out subpoenas. Each of us had to reveal what strains we had. They went through those. And once they discovered that a laboratory had the Ames strain, then that laboratory had to make a copy of each one of their cultures, and then send it into the FBI.

So the repository of Ames strains was being assembled. And this was not something that happened in 30 days. This took months and months to identify the laboratories and then to assemble those. Those samples were sent in duplicate to the FBI laboratory that was then at USAMRIID. So they were sent to USAMRIID, and then they were anonymized by giving them a number that would make them indistinguishable to anyone from the outside.

One sample was sent to my laboratory in Flagstaff, Ariz. The other sample was then kept at USAMRIID. So we had a duplicate of each one of these. When those samples came to my laboratory, I would then extract the DNA from those samples, and we would run the Ames strain test to see if it was Ames strain or something else.

As it turned out, maybe 5 percent of the samples that got sent to us from around the United States weren’t the Ames strain. The people had not kept careful track of what they had, so those then were eliminated from the investigation while the Ames strain continued to be investigated. …

Eventually 1,070 samples were gathered from multiple labs. What was the process? Was it clear enough? Some people complain about the fact that it was voluntary. Some people felt that the submissions were not all consistently done in the same way, which also caused complications. What’s your overview of this huge process?

It was a massive undertaking to go out and get all these samples. The people who actually were sending the evidence in weren’t necessarily the perpetrators. It’s quite possible that you’d have a person doing the sampling from a laboratory, and there was another person in that laboratory who had actually been the perpetrator. So given the complexity of it, to send somebody out to witness taking all these samples was pretty much impossible, I would say. It was probably done about as well as it could be done at the time.

The process was to — well, first off, they had to identify where the Ames strain was. In 2001 we were all regulated, but we weren’t required to submit a list of our strains. We weren’t even required to keep track of how many vials of anthrax we had in our laboratory in 2001. The rules have dramatically changed now, and if you had asked me right now how many strains I had or what their names were, I could tell you very quickly going back to our inventory.

But in 2001 that was a chore. We had to pull together a list. We had to come up with how many different strains we had, how many different samples we had. And all that had to be supplied to the FBI under subpoena.

So while I was processing the samples for the FBI, I was being subpoenaed as well. So not only was I working on the investigation, but I was being investigated, and that continued for quite sometime. In the early days after the anthrax letters went out, it was very intense, and there would be two different themes in my laboratory, the FBI scientist working with the evidence, and then the FBI investigators would come in and say: “Where were you, Dr. Keim, on a certain date? Do you have an alibi?” And of course Flagstaff, Ariz., is 3,000 miles from Washington, D.C., so I had pretty good alibis.

But that dichotomy continued as a repository was being built. And in many ways I knew what was going on in the labs that … because I was being investigated. I would get the subpoena, and I would read those, and I would have to go in and make the samples.

So the samples that I prepared from my laboratory got sent to Fort Detrick in Maryland. And then a duplicate, half of those samples got sent back to my laboratory after they’d been relabeled. Of course I didn’t know which ones they were because all the tubes look the same. But I was actually processing and analyzing the samples that I had submitted under subpoena the days before.

The process in those early days was very clearly defined in the subpoena, so those of us who did it knew that we were supposed to collect it a certain way. No one was watching us. It wasn’t required that anyone was watching us, so it was obviously possible for somebody not to follow the directions. And that’s a weakness in the repository, and it’s one that I think if we had to do it all over again, we would do it differently.

Some people will say it was a flawed process … because if a scientist who was guilty of the crime had the ability to control the sample, he’s not going to send it in, or he’s going to destroy it, and there could be possibly no way that you’d ever discover that that in fact had taken place. …

Perhaps that’s true. I mean, if a sample is kind of obscure and not being well tracked within a laboratory setting, you might be able to destroy it, and then it would be unknown. But if it’s a high-profile sample — for example, a flask with 1,000 milliliters that’s being routinely used for lots of animal challenge experiments — destroying that would be noticed, and it would of course draw attention to the perpetrator. So that’s the other side of that.

To say that it was a perfect process, it wasn’t. What we were looking for early on in the investigation changed. Later on, when the morphological variances were being pursued and more intensely, then the purpose of those samples changed. So the original subpoenas were done about as well as they could have been done without knowing what we were going to do with them.

And basically the FBI needed some luck if this process was going to show up.

Well, we thought we were doing one thing, and we ended up doing another thing in this process. So the genetic fingerprints that eventually led to RMR-1029, yeah, we got a little bit lucky, and they led back to something that matched the letters.

I hope people don’t think that this DNA fingerprinting that we’re doing is anywhere as strong as the DNA fingerprinting that’s used for human identification. You cannot use this type of DNA fingerprinting to convict someone. You’re going to have to have additional evidence.


Your DNA that you have in your body is really unique. It’s so different. It’s so easy to distinguish it from anyone else — other than an identical twin — that we go into court with human identification, and we can get a conviction with no other evidence.

In the case of this particular profile that we have from the anthrax letters, that was shared by more than one sample. It wasn’t just the RMR-1029. There were seven other flasks, seven other samples that had that.

There were hundreds of people perhaps who had access to that, so all that really did, all the DNA profile did, was narrow a very large bull’s-eye from hundreds or thousands of people down to a much smaller number. At that point then, the FBI had to use its traditional, classical forensic investigative methods to find out who could have sent that letter.

How does it get narrowed down to this RMR-1029 flask? What’s the process? And were you involved in that in any way?

A little bit. Our role was that we were handling the repository, so that was part of it. So we were extracting the DNA. And then the other side of it was TIGR, the team, Jacques Revel and Claire Fraser, were doing the whole genome sequencing. They have no capability to handle dangerous anthrax, so we would do the DNA extractions. We would send them to Maryland, also just north of Washington, where they would run their big DNA sequencers, and they would determine a whole genome sequence of this, which led to the morph identification signatures.

When do you find out that there seems to be eight samples that are showing more and more morphs, and that’s leading them back to one place?

I didn’t have all the details of that until after Bruce Ivins committed suicide. Personally, I was kept out of the loop on exactly all that was happening there. In fact, none of the scientists that were outside of the FBI knew about the match, I think, until after we were released.

How did this process lead to one flask?

So there were these strange little morphological differences in the letters. Those then were analyzed using a highly precise method of whole genome DNA sequencing. Once that was done, it was discovered by Jacques Revel and his team at TIGR that these little morphs, these little rare variances, actually had differences in their DNA. Those differences were unique to, in fact, the spores.

So by using diagnostic tests to track those rare morphologies, tests were developed where you could screen different samples for the presence or absence of those rare morphological differences. Now we’re looking at the DNA. We’re not looking at these on a plate. We’re looking at the DNA.

In the end, the FBI and the scientists involved with the FBI decided to focus on four different ones, because they had very good tests for them. And these are four subtle differences that were found in the anthrax letters that aren’t normally seen in anthrax. Those tests were then run across the entire repository of 1,070 samples. They found that there were only seven samples in there that matched up with the anthrax letters, and these were seven samples that seemed to trace back based upon notebooks and testimony that had been given by scientists.

These seven samples trace back to one sampled called RMR-1029, but it didn’t have the morphs. So the question was, how can all of these things, all these samples, come out of RMR-1029, and yet RMR-1029 itself didn’t have the morphs? That was the big question. …

I thought there were eight samples.

The eighth sample was RMR-1029. There was RMR-1029 plus seven samples, so the original result apparently was seven without RMR-1029. It was only the eighth that was finally identified. …

Why did we think that the 1029 did not have the morphs?

The FBI didn’t understand this at all. My understanding is they went in and they sampled RMR-1029 themselves and basically took possession of it away from the laboratory, which was Bruce Ivins’ laboratory at that time. They took RMR-1029, and then they analyzed it themselves. When they analyzed it, they found the morphs.

So in the repository, the RMR-1029 tube didn’t have the morphs, but the RMR-1029 in the laboratory that was controlled by Bruce Ivins did.

So explain why all sorts of bells and whistles went off.

At that point they began to look back to see how the sample in the repository had been collected from RMR-1029. What they knew was that Bruce Ivins had collected that sample; he had sent it in. So they instantly began to wonder if this wasn’t an attempt to obscure the fact that RMR-1029 was the source of the anthrax letters.

In doing so, they began to review their records, and they realized that Bruce had actually submitted a sample from RMR-1029 twice. The first time he had done it using a procedure that they didn’t accept, so they made him go back and do it again. But they had gone ahead and sent that sample to my laboratory, and it had been put into my collection as part of my part of the FBI repository. So it was still in my laboratory.

Now, I didn’t know this. … It was an anonymous sample that was numbered in a way. But they called up and said, “Do you have such-and-such a sample?” And I of course went back and looked and said: “Well, of course we do. You sent it to us, and we saved everything you sent to us.”

They said, “Could we get it back?” I said: “Well, of course. It’s your sample.” So they said, “OK.” So they actually sent an FBI agent out who collected the sample and took it back to USAMRIID. I had no idea why they were doing this, not until August 2008, when basically they told me and other scientists the bottom line on the entire investigation.

So it was the FBI that figured out that this other sample still existed. So in 2006 when they come to get it, you’re basically just —

Yeah, I’ve been portrayed as a real hero for saving the sample, but it’s just what we do. We save samples; we collect them; we keep track of them. They’re of course very well protected and locked. So the FBI, they actually hadn’t seemed to realize that that sample still existed, because there are two copies of every sample. One is kept at USAMRIID, and one was sent to me. They destroyed the sample that was to be kept at USAMRIID because then they asked Bruce for a second sampling of RMR-1029, which also came to my laboratory.

… What were the differences within this first sample that you supplied?

… RMR-1029 is a flask of spores that was very large. It was very large because it had to be used for a long period of time. When you’re doing vaccine development, you have to vaccinate the animals, and then you have to challenge them with virulent anthrax to see if they die or not. Bruce had developed this flask which contained a very large number of spores that could be used year after year after year and remain unchanged. So it was a large reagent. …

It was a very unique set of spores. It was actually a composite of multiple batches of spores that had been produced in Dugway, Utah, where many of the large spore-production facilities are for the U.S. Army, and then also spores that Bruce had produced himself there at USAMRIID. It was a big mixture of these things that had been thrown back together.

So obviously, that’s where the morphs came from. The morphs must have arisen in this production process. And because RMR is really kind of a unique type of a sample, or flask, it had a unique composition that isn’t seen in other samples. …

Did the FBI ever talk to you about this and about the suspicions that were raised because of the fact that the 1029 originally in the second sample that Ivins sent to them proved not to have the morphology?

I knew very little about this, and certainly nothing about Bruce until really very near to his death. I learned that the FBI was closing in on Bruce just a couple of months before his death, when they began to try to reconstruct the timeline for when Bruce might have figured out what was happening. I think it was in April or May of 2008 when I met with about five FBI agents and five Justice Department lawyers in a hotel near Dulles Airport, and for an hour they quizzed me about when I talked in public about our DNA fingerprinting methods, when we published, when people outside of the investigation might know what we’re doing, when people inside the investigation might know what we’re doing.

In an hour they took a break, and they did a sidebar out in the hallway. Left a couple of FBI agents with me, and we chatted. Then the lawyers came back in, and they put in front of [me] the e-mails that had been exchanged between myself and Bruce Ivins, and also between Bruce and other people, and said: “What was Bruce thinking when he said this? And what were you thinking? What would he have known?”

And that’s when it became clear to me that they were investigating Bruce Ivins. But that was just a couple of months before his death.

… What were your thoughts when you looked at the e-mails?

I didn’t see a smoking gun there. Of course, they were hoping that I would see something that would prove that Bruce had sent the letters. I looked at it, and I go: “Well, Bruce was very curious about the Ames strain. He was very curious about lots of aspects of the investigation. But we all were. Everybody was curious about it. We were anthrax researchers, and this was the biggest thing that ever happened.”

So I didn’t see anything there that was out of the normal in those e-mails. …

What did it seem like they were trying to understand better?

During that early 2002, Bruce and I and many other people in the anthrax research community were communicating about the letters and how the spores were made and what type of DNA fingerprinting methods were available. So the FBI was trying to understand if something that I had said, or something that Bruce might have asked, might have actually tipped him or other people off to the fact that we were starting to get deeper into the genetics of Bacillus anthracis, and that this would, in fact, lead to a way that we could identify the source of the letters. …

Looking back at it now, how important do you think the second submission was to the FBI in their analysis of guilt or innocence of Bruce Ivins?

It’s beyond my purview to [know] how important that was versus everything else.

But they certainly were interested in following through on this?

I’ve heard them talk about this has been part of the circumstantial case against him.

Were there, as far as you know, any other submissions by Bruce?

Oh, hundreds I would suspect. You know, USAMRIID had the largest collection of Ames strain of any facility in the world. They certainly submitted hundreds. So Bruce must have submitted hundreds of different submissions to the repository. But again, I don’t know the exact number. …

Speaking again as an expert on this, … his lawyer says that Bruce in his interview said that he had pinpricked the second submission, and that’s why it didn’t show the morphology. He was not trying to hide anything, but that he had done it that way; there was no written protocol and that he misunderstood. What’s your take on that?

Certainly if he had gone in and picked a single colony, as we would say, he would not have submitted the four minor morphs as part of that sample. That’s a very common way to submit samples and the culture, and it’s certainly the way we would have done it.

But the subpoena was very specific on how you were supposed to sample. You were supposed to take a pretty sizable sampling of multiple colonies. In this case it was liquid, so it should have been using a loop. He should have been able to sample the morphs if he’d followed the subpoenaed instructions.

And based upon what you’re telling me, his lawyer says he did it this other way. That wasn’t what the subpoena said.

And you got that subpoena.

I got that subpoena, yeah.

And that subpoena was very clear.

It was a little bit of a surprise to me that they wanted us to sample it in this way, because again, it isn’t standard microbiological technique to do it this way. But when you read it, you knew what they wanted. It wasn’t confusing; it was just a surprise that they wanted us to do it that way. …

What’s your belief at this point of the possibilities that Bruce Ivins did or did not do that?

The Bruce Ivins I knew  — I never suspected that he sent the letters. I thought he was a great guy; he was a good colleague; he was a smart researcher; he worked hard. I would not have picked him out of a lineup.

Since then, many facts have come out, or a lot of information has come out about Bruce that I didn’t know. One of the most damning reports I read is the psych report. There were things about Bruce that were dark, and there were things that I didn’t know about. You know, I have to say that it’s a possibility that Bruce did this based upon those things I read in the psych report. But did he do it? It’s hard to be absolutely sure about that. …

His defenders will also say things like he didn’t have the ability, the knowledge to work with the dry stuff.; he didn’t have the equipment; he didn’t have the time. Speaking as an expert in anthrax, what’s your evaluation of his capabilities?

I don’t think that these spores are all that special. I think that it would be possible to grow these spores up in a normal BSL-3 [biosafety level 3] laboratory like Bruce had access to. I think that it would be possible to produce these spores in that type of a setting.

One of the things the other people point to the fact is, if he dried it in a lyophilizer, you would have found spores. No spores were found — no spores on his body, no spores in his car, no spores in his house, no spores in the lyophilizer; no DNA evidence even if the spores were killed, no DNA evidence within the lyophilizer.

I know that we have equipment like this. We don’t have spores in ours. You do not use equipment like this and end up with spores everywhere. If you think you do, you have very thorough decontamination.

One of the cleanest places in the world is inside my anthrax laboratory. If you go in there and look for spores, you won’t find them, because we absolutely keep those places clean. People who work with anthrax know how to clean up things, so I don’t find that convincing at all. …

What was the mood at the press conference? What was your role in it? What happened?

Well, first off, it was twofold. In the morning we met with science writers to explain the science. There were scientists there, as well as FBI agents and U.S. attorneys, people from the U.S. Attorney’s Office.

So in the morning we met with science writers, and the scientists like myself, Jacques Revel, Jim Burns, Claire Fraser, we explained what we’d done and how the science worked. Very civil. There was very detailed questions. Then we had lunch. Then we came back for the afternoon, and then it was the entire media, and there were maybe 20 or 30 representatives. And it was anything but civil. It was just crazy.

The science became very much back burner. The scientists said almost nothing. It was almost all interactions between the reporters and FBI agents and the U.S. Attorney’s Office. Other than just a little bit of a superficial explanation of what the science was that led us to this point, we were irrelevant. The media didn’t care about us anymore. They were after all sorts of other things.

Some people say, looking back at that, that DOJ folks oversold the science too much at that, trying to bring a close to this case. What’s your take on that?

Was the science oversold? Maybe. Or at least misunderstood. Those of us who were involved in the science knew what it said and what it didn’t say. And in fact I spent a lot of time trying to explain to the media what it said and what it didn’t say. I was released from my nondisclosure agreement with the FBI for that press conference, so suddenly, after seven or eight years, I was able to talk to the media without getting in trouble, or without the chance of going to jail over it.

I took that mission very seriously, and I talked to people and tried to explain to them what it said and what it didn’t say over and over again. … I said: “Look, you’ve got to go back and figure out what the circumstantial evidence is that’s pointing toward Bruce Ivins, because the science did not point to Bruce Ivins; it pointed to RMR-1029. And Bruce’s connection to RMR-1029 was something that was established by totally different methods.” …

Were you surprised? What was the bigger meaning of it?

Of course I didn’t know that he was suicidal or that this was the way it was going to go. But I knew he was in the crosshairs. I knew that he was under intense pressure. I can only imagine the type of pressure he was under. So when I heard that he had died, you know, I didn’t expect it, but it wasn’t a big shock either, because I knew that he had to be under intense pressure. …

Some of his friends kind of say that he was hounded to death by the FBI. What’s your take on that?

I guess the answer to that question is yes, he was hounded to death. Whether he was innocent or guilty, I think that pressure is what pushed him over the top. Could the FBI have done it differently? I don’t know. I mean, I’ve talked to the FBI and I’ve talked to the U.S. attorneys, and they went about the investigation the way they normally do it.

I’m not an investigator, so their methods certainly probably include putting pressure on someone in order to get a confession. If that’s what was happening, it didn’t work to get the confession, but it was a lot of stress.

What’s your take on motive? …

If his motive was to change the direction of U.S. science, U.S. science changed dramatically. We’re talking about billions and billions of dollars now are being spent in this area of biodefense research. And people like myself and programs, research programs like myself have expanded dramatically. My research laboratory is 10 times larger than it was pre-anthrax letter attacks.

And Bruce’s research program was reinvigorated, because suddenly we were very interested in anthrax letters. Is that motive? Well, I wouldn’t have sent the anthrax letters for that outcome. I would have been very happy with my life before the anthrax letters. So if that was the intent, it worked.

The NAS [National Academy of Sciences] report, were you surprised by the findings? What do you think the most important things were that they added to the discussion of the investigation?

The NAS report was actually kind of a soft endorsement of the science, from my point of view. The science was so revolutionary. I mean, to go in and try to identify the source of a bath of spores based upon its individual components had never been done before. And that revolutionary science was endorsed by the committee.

They came back and said you need to be very careful about the confidence that you have in these results, and try to establish confidence estimates for matches. And all that was very interesting scientific discussion. But the overall approach was endorsed. …

They did say, though, that it was not possible to reach a definitive conclusion about the origins of the anthrax and the mailings. What did they mean?

Unlike human DNA fingerprinting, we just don’t have that level of confidence in this approach. I think part of their caution was in fact the way the samples were collected, which was not as controlled as we’d like. The theory for how these different morphologies ended up in those spores is not well established either. I mean, we haven’t gone back and reconstructed these types of experiments.

In other words, how do these morphs arise? How do they end up inside of a batch of spores? Those are all studies that they recommended be done in the future. Are we ever going to do them? Well, if there isn’t another set of anthrax letters, probably not. The motivation here for us or the initiative for us to continue doing this work is going to be much less intense than it was several years ago.

We are very interested in understanding these things, so we will continue working on it, but not at the same speed or intensity that we were before.

They also say the mutations might have arisen by parallel evolution rather than by derivation from RMR-1029.

… There are 1,070 samples, and all four morphs only occurred in things that came from 1029, so the best argument against the parallel evolution is in fact [in] the entire United States, with over 1,000 samples, we only saw it in those eight samples, and they all were related to RMR-1029. So could parallel evolution occur? Yes, it definitely could. And that’s something that is going to be explored in the future with scientific studies. Did we see parallel evolution in these set of samples, in all the Ames strains from around the world? No.

They also said that the FBI and DOJ possibly overemphasized what the evidence meant. … It was overstated that a disputed sample submitted by the suspect had not come from the proper source. What’s your take on that?

It hadn’t been sampled correctly. I mean, that part is obvious, because RMR-1029 had those minor morphs in it, and they showed that by repeated sampling over and over again. So the way it was sampled the second time was not done properly. Whether it came from RMR-1029 and it was only a single colony pick, that’s a possibility. I mean, it could have come from RMR-1029, but by a different procedure than had been used previously. …

Your evaluation overall of the FBI’s investigation? It took eight years. Do you think it was successful in the end? …

I’m very much an amateur in this. Of course, this is the only investigation of this type I’ve ever been involved with.

I thought that there would be an indictment very early on within weeks or months of the anthrax letters. I thought that the number of suspects would be very small and that they would be able to actually make an indictment very quickly.

Then of course I became pretty discouraged that we would ever find out. And then I began thinking the only way that we would ever get an indictment was like in the Unabomber, when someone came in and identified something.

So did they do a good job? We did some amazing science here. I think we tracked it back as close as we could to the source with the science. After that it would have been nice, of course, if they would have had definitive [knowledge] of who did it. And I think that that is lacking here. It’s a circumstantial case, and how strong that is we’ll never really be sure, because we don’t have a jury listening to this evidence, do we?

Any surprises? What’s the biggest surprise out of all this time you were involved with this case?

The biggest single surprise? Two big surprises was the Ames strain — in those early days, that was a shock and a surprise. Then the second thing was when they told me that it was Bruce that they were investigating. That was a real surprise, too.

More Stories

‘It’s All We Could Do: Cry’: A Honduran Mother and Daughter Describe Their Separation Under U.S. 'Zero Tolerance' Immigration Policy
A Honduran mother and her young daughter, separated under Trump administration immigration policies, share their stories in a new FRONTLINE documentary, ‘After Zero Tolerance.’
December 6, 2022
Kherson After Liberation: Co-producer of ‘Putin’s Attack on Ukraine’ Documentary Describes Visit
A Ukrainian filmmaker and journalist who was on one of the first trains traveling to the newly liberated city talks to FRONTLINE about the damage he saw in Kherson after eight months of Russian occupation.
November 29, 2022
As Donald Trump Announces His 2024 Run, a Look Back at His Presidency and Impact
FRONTLINE has built a unique public record, in documentary format, of the former president’s impact on American life, politics and democracy — and his previous battles with a special counsel and the Department of Justice.
November 16, 2022
How American Politics Reached This Fraught Moment: 12 Documentaries to Watch Ahead of the Midterms
As a divided America prepares to vote and fears of political violence continue, these FRONTLINE documentaries show how U.S. politics reached this moment.
November 4, 2022