I’ve decided to liberate as much of my work as I can, exploiting the advantages that having an institutional repository has. So why not start with paper #1.
Back in 1985 I met Daniel Oliva, a medic working at the Institute of Cell Biology (now known as the “De Robertis Institute”). It was a great opportunity and the lab had a few rules: you don’t get to do anything until you know what you are doing and understand why. I was even given a scientific paper to read about how to wash lab glassware before I was allowed near the sink (really). I washed lots of dishes, sterilised a lot of glassware and participated in the daily lab life.
One day we came across a 1982 paper by Lilly Yeh Jan and Yuh Nan Jan reporting than an antibody raised against horseradish peroxidase could be used to label neurons in insect brains. We didn’t have a lot of money in the lab, but we did have those antibodies – we used to raise those ourselves because we used them in an antibody labelling technique called PAP (for peroxidase-anti-peroxidase). We immediately wondered whether that might be the case for rat brains, which was the animal we were using at the time. Back then there weren’t dozens of catalogues selling you the antibody of your choice, and even it that had been the case we would not have been able to afford them. We looked puzzled at Lily Ye Jan’s paper – surely she must have tried this in mammals – imagine the power of being able to label neurons in such a way! But if she had tested this, and failed, why would she not report it?
So was born my “first” independent experiment. Try to see if we could get the same labelling in mammals. We had, after all, plenty of rat brain tissue to test it in. But as the good training ground that lab was there were a few hurdles I had to overcome. The first was that I needed to replicate Jan’s experiments. I knew (because I was studying biology) that there was a group at the National Atomic Centre CNEA) working with Drosophila. I managed to get a hold of the group and off I went to collect my first fruit fly vials.
The day I went to collect the flies is burnt in my memory. The bus took me to a few blocks away from where the lab was, and to get there I had to walk past the ESMA building, infamously known for being a place where the militaries had performed human torture only a few years back during the dictatorship. I had driven before past that building, had gone past it on the bus. But the slow walk to the entrance of the CNEA was just, well, different. I can’t remember this project without being reminded of that.
I eventually made it back to the lab at the med school and pulled the fly vials out of my pockets. I was to embed them in paraplast and cut sections in the microtome, techniques I had never really done before on my own. After reading the manuals and being tested on the ins and outs of how the microtome worked, how knives were sharpened, the chemistry of the process, and a lot of other stuff, I was given the go-ahead.
My first set of sections had a lot of little tiny holes in them – something that sometimes happens when you don’t do the embedding well and get bubbles in it, and so it was dismissed as such. I tried and tried changing the embedding conditions to no avail – those bubbles persisted. I thought it might be that the paraffin was not getting through the fly’s exoskeleton. We were all puzzled. Until I realised those were little holes from fruit fly puppae that had hatched. Duh! Problem solved, move on. (think #overlyhonest here)
When I finally got my first few good sections we labelled them with the antibody and voila, first successful replication. Of course this was repeated until we were certain, and then I tried the antibody on rat tissue. It didn’t work. I had done all the controls, so I knew it wasn’t a technical error, and after a few repeats we were convinced of the results.
We thought we should publish this negative result, after all, we most surely were not the only ones who would think of this, and the information would be useful to others, if nothing else so they wouldn’t waste time and money chasing immuocytochemical ghosts. I wrote the paper with my friend and colleague Su Peressini, and sent it as a short communication to a small journal published by the Institute. The paper got accepted and published in 1987.
Not long after I happened to be visiting friends in San Jose, CA and decided to visit Lily Yeh Jan at UCSF. I asked her about whether she had tried it on rats and this is how I remember the conversation – bear with me – this is very old memory! As far as I recall she said she knew the labelling didn’t work in vertebrates. I then asked her how she came to discover the antibody worked in insects. The way I recall it (see memory disclaimer above), she said that they were using those antobodies in a way similar to us, and that one day a worker in her lab was puzzled by finding this staining in neurons. Tracking down their steps they realised they had mistakenly omitted a step – that this fantastic finding was the result of a “mistake” made when following the steps on the protocol.
Many years later I received an email from Dr Jan, asking me for permission to use an image from one of the papers I published as a postdoc (which of course I granted). For some reason that request made me smile, knowing the memory of our previous encounter would be one sided.
The protein that the antibody recognised has since been uncovered. At about the same time, the journal where I published the findings was discontinued. Just over 25 years later I somehow take pride in having my first paper be about a replication and a negative result. Now, just trying to get it into the repository.
- Jan, L. Y., & Jan, Y. N. (1982). Antibodies to horseradish peroxidase as specific neuronal markers in Drosophila and in grasshopper embryos. Proceedings of the National Academy of Sciences, 79(8), 2700–2704.
- Kubke MF, Peressini, SM and Pecci Saavedra (1987) Lack of a peculiar antigen in rat. Comunicaciones Biologicas, 5(4) 391-395
(and HT to Bjoern Brembs who encouraged me to write the story)
This work, unless otherwise expressly stated, is licensed under a Creative Commons Attribution 3.0 Unported License.