Life Science Tools of the Trade

It is no longer a secret that synthetic chemical pesticides, insecticides, fungicides, fumigants have a definitive link to chronic dis-orders including various forms of cancer [1].

The other day while browsing, I came across a very interesting paper by Dr. Ames (famous for Ames test for carcinogenesis) work published in 1990 on Dietary pesticides [2]. According to this paper; 99.9% of the pesticides by weight in American diets come directly from plants – that produces them naturally to defend themselves. Out of which about 27 are found to be carcinogens on rodents!! The authors estimated that an average Americans consumes about 1.5g of natural pesticides every day which is 10,000 times that of the artificial pesticide they consume. This is a startling revelation. The paper further argues that cooking food causes production of few more carcinogenic materials starting from polycyclic Hydrocarbons, heterocyclic Amines, furfurals, Nitrosamines, Notoaromatics. All these really sound scary isn’t it? The paper tells us that nutritious vegetables such as Broccoli, cabbage, cauliflowers and other crucifers can be carcinogenic.

I decided to take a dig on few phytochemicals that have been labeled as carcinogens in this paper:

Glucosinolates (Found mostly in cruciferous vegetables etc.)

Chlorogenic acid(Found in Apricot, Cherry, peach, plum etc.)

Caffeine(Found in Apple, Carrot, Celery, Eggplant etc.)

And majority of Phenolics(Found in Coffee, lettuce, grapes etc.)

From this list, it mainly appears that the vegetables and fruits we considered most beneficial are labeled as potent carcinogens in rodents. There are a very large number of papers published over a period of time, that vows on beneficial effects of the above mentioned fruits and vegetables on cancer. For example:

“Epidemiological studies suggest that brassica vegetables are protective against cancers of the lungs and alimentary tract. Cruciferous vegetables are the dietary source of glucosinolates, a large group of sulfur-containing glucosides. These compounds remain intact unless brought into contact with the enzyme myrosinase by pests, food processing, or chewing. Myrosinase releases glucose and breakdown products, including isothiocyanates….” [3]

“Epidemiological studies indicate that human exposure to isothiocyanates and indoles through cruciferous vegetable consumption may decrease cancer risk, but the protective effects may be influenced by individual genetic variation (polymorphisms) in the metabolism and elimination of isothiocyanates from the body…” [4]

“Plant-derived phenolic compounds manifest many beneficial effects and can potentially inhibit several stages of carcinogenesis in vivo. In this study, we investigated the efficacy of several plant-derived phenolics, including caffeic acid phenethyl ester (CAPE), curcumin, quercetin and rutin, for the prevention of tumors in C57BL/6J-Min/+ (Min/+) mice…” [5]

Chlorogenic acid, caffeine: “Many research investigations, epidemiological studies, and meta-analyses regarding coffee consumption revealed its inverse correlation with that of diabetes mellitus, various cancer lines, Parkinsonism, and Alzheimer’s disease…” [6]

So, these studies clearly contradict what Ames said in his classic paper. What Ames paper did not touch upon was the bioavailability of the carcinogenic materials derived from natural resources. Moreover, Phytochemicals are not assimilated in  the body as pure chemicals. However, somewhere in the middle of the paper, the authors coyly agree that vegetable consumptions do reduce the risk of cancer. When majority of people just read the title and abstract, will definitely get a wrong signal from this paper. I would like to hear more from the readers on this issue.

References:

1. http://www.ncbi.nlm.nih.gov/pubmed/22571220
2. http://www.pnas.org/content/87/19/7777.full.pdf+html
3. http://www.ncbi.nlm.nih.gov/pubmed/11887749
4. http://www.ncbi.nlm.nih.gov/pubmed/17317210
5. http://carcin.oxfordjournals.org/content/21/5/921.short
6. http://www.ncbi.nlm.nih.gov/pubmed/21432699

This topic has been simmering in the background for a while. I can see that it could stay there for quite a while, with me adding occasionally to each category, but I think readers (you do exist, right?) might enjoy adding their own 2 cents worth in the comments.

You shouldn’t read too much into the longer list of things I hate. It comes with age (see previous post about blots). Nice things just don’t give the brain the same jolt that bad things do. Perhaps that’s an evolutionary thing – we survive because we remember the scary/dangerous things.

Alright, that called for a Google – and it seems that there are different opinions on the subject. But from PsychCentral:

“…The speculation is that we process memory in order to solve problems. And things we should learn from, things that are particularly important or that have strong emotions tied to them, may be things that are going to be important in the future. If you present stimuli with a strong negative emotional component, the memories do seem to be more easily retrieved than neutral stimuli or even those that are somewhat positive…”

Phew, maybe I’m not just a negative, miserable git after all.

Things I Love

• Wife, kids, dog and poor budgie with tumor.
• The rush of finding something really new and exciting in the lab. Doesn’t happen often enough (of course), but it’s very cool when it happens. Probably the biggest was when I was a post-doc and discovered that poxviruses encode an interferon-gamma binding protein (it helps block the host’s immune response to the virus).
• The relief at getting a paper published -  this tends to fade all too quickly.
• BLAST searches. It still amazes me that I can send a query from my desk to where ever NCBI’s server reside (still Maryland?) and get an answer back in a matter of seconds, never mind that BLAST can compare my sequence to 140 million other sequences in this time too. It’s the movement of the info that’s mind-boggling -  I wish I understood more physics!
• Leonard Cohen.
• My iPad.

Things I hate

• Anthropomorphisms. It’s not just the blatant ones, it’s the ones that suggest that “the flu virus changes its antigens because it wants to avoid the immune system” that bug me.
• Scientists who insist on describing two proteins as having 50%  homology.
• People who don’t vaccinate their kids.
• Homeopathy.
• Abuse of statistics and inadequate descriptions when statistics are quoted in the media.
• The decline of investigative journalism. It seems to me that “reporting” too often just gives someone a platform to spout their point of view without any serious attempt to question it. Then there’s the interviewers that follow a set list of questions disregarding any answers.
• Evolution naysayers. Checkout the Miss USA contestants answers to “Should evolution be taught in schools?” if you dare.
• Web police who tell me how our departmental web site should be formatted, with information hidden behind multiple clicks.
• Poor web sites.
• Scientists who don’t reference software because “it’s free”.

Artist Guido Rosso: Academy of Art University

My first blog for SAB! I want to talk about transitioning from low-throughput to high-throughput. Mostly I’m going to talk about high-content imaging analysis, but let’s start with a little historical context.
My experience in science began from the early days of PCR circa 1993. Back then, it was very low-throughput. Unlike most of today’s grad students, I remember racking individual tips by hand and autoclaving them to use with manual single-channel pipettes. Lately I’m going through several boxes of pre-racked sterile tips with digital multichannel repeat pipettors, and excited about the prospect of soon using a robot to set up multiple 384 well plates at a time.

I’m really excited about finally going from doing immunofluorescence on acid-washed coverslips I prepared and separated manually using forceps for the last ten years or so, to setting up optical 384 well plates automatically using a robot (a robot!) and programming all the staining steps to be done using a plate washer.

Because this transition is a little scary, I made a list of pros and cons regarding scaling up.

What’s really going to change? The workflow is drastically different in terms of the potential bottlenecks.

Pros:

-Lots of data in little time!

-Mostly hands-off, go think big thoughts while robots do the tedious part!

Cons:

-One big machine shared by many people= signing up for shared equipment still not my favorite thing. Hard enough scheduling cells, transfections, time points after drug treatments, etc. without having to work around other people’s equally unpredictable schedules.

-Data transfer might even take as long as collection. Heavy dependence on computers – analysis can be the slowest step.

-Big specialized machines and few alternatives = if the robots are down, your experiments are on hold and your samples might end up in the trash if the delay is longer than fresh biology will allow.

Things learned along the way:

1. Sample prep and controls are still 99% of the work (but we knew that already).

2. How to interpret and present the data.

For example, how much agreement should we expect between a cell population distibution vs. whatever western blots are measuring?

We tend to treat western blots as if they report an average protein levels, but really it’s more like a weighted sum: the highest expressing cells in a mixed population will dominate the signal. Really, it’s not even clear whether the same cells are expressing two or more proteins of interest at the same time, or if it could be separate sub-populations of cells, but westerns are often treated as a proxy for this assumption. Combined with immunoprecipitations, it’s a reasonable interpretation. But what if you can’t IP a complex?

Point being, looking at population distributions on scatter plots with other variables really exposes the strength of understanding the difference between a possible correlation (appearance of two subsets of cells) vs. a potentially direct dependence (one subset of double positive cells).

It’s useful to think about different ways to define a change in signal in a population, e.g. threshold cutoff (for a clearly separate sub-population) or average intensity (only makes sense if the staining is uniform, which it almost never is).

Pros:

Some people who were uncomfortable looking at representative images will (sometimes) be more comfortable with graphs.

For adherent cells, this is much better than flow cytometry! The possibility to reanalyze old images is wonderful with such large datasets. It’s great for discovery – easy to examine variables like cell cycle, protein localization, and morphology.

Cons:

Telling the computer how to measure what I see can be really challenging. Exporting the data table in a consistent format isn’t always possible. Conversions are necessary.

Non-microscopists using high-throughput microscopes tend to avoid even looking at the primary data. This is a sad thing for discovery science, since the real power of microscopy is in seeing things you weren’t looking for.

Arguing about the mathematical representation of biology with the people who write & sell the software. E.g. the “it’s your sample” excuse when one commercial algorithm can’t segment my nuclei (but here’s the thing, I know of other ones that can!).