Supplemental Material Box 1: Victor Zhdanov’s discovery of mitochondrial viruses.
In the early 1970s, an esteemed Soviet virologist, Victor Zhdanov, reported that the mitochondria can, in some cases, serve as an alternative site of viral replication
(1-4). Owing to the enormous political influence Victor Zhdanov had in the country, it was difficult to contradict his observations or to get funding for an independent assessment. However, studies devoted to the analysis of mitochondrial DNA constituted a completely separate, independent and “remote” field. With the discovery that the mitochondrial genetic code is different from that of the nucleus (6), questions were raised and it became clear that the viral genomes Zhdanov was working with were not compatible with the mitochondrial machinery.
Supplemental material Box 2: New technologies challenge old conclusions.
Let us assume that one is interested in knowing which amino acids of a mammalian or viral protein are essential for its activity. For years, researchers introduced amino acid substitutions into the protein and expressed the wild type and mutant forms in bacteria. Obviously, the scientists knew that bacterial expression system only partially mimics mammalian cells. However, in mammalian cells, the expression levels of wild type and mutant forms of the protein could be dramatically different (6). As a result, one could not compare their activities and conclude that the difference resulted from the alteration in the amino acid sequence, and not from the different quantities of the two proteins. Recently, it became possible to introduce mutations into the gene while maintaining the equivalent expression level. This method, which works at least for some proteins, is based on the discovery that reduced expression levels occur because a mutation disrupts the secondary RNA structure (6). In the same paper the authors demonstrate that by designing mutations in silico, which would alter the amino acid sequence while preserving secondary RNA structure, one preserves the expression level. This method should allow for a reexamination of many conclusions drawn from bacterial expression systems.
Supplemental material Box3: Hormone-independent activation of steroid hormones.
Hormone-binding proteins were initially identified in hormone-sensitive cells, and given the names of the hormone receptors they bound (e.g. progesterone receptor, estrogen receptor, etc). These results led to the beginning of the new field of molecular endocrinology, which has grown to an important part of today’s medical science. Medical textbooks all over the world stated that the receptor is not active in the absence of the hormone. They stated that hormone binding activates the receptor, which in turn activates or suppresses the expression of downstream genes leading to e.g., lactation or oncogenesis.
This mechanism of action did not explain, however, why some steroid receptors are present in tissues where the hormone is absent (e.g., progesterone receptors are abundant in the brain, whereas progesterone is believed to be absent). Yet this observation was either ignored or was shrugged off as possibly “nature’s whim.” As it was understood that hormone receptors are activated by their corresponding hormones, nobody asked the question: “Could there be a different mechanism of receptor activation besides hormone binding?” Nobody, that is, until the work of Bert W. O’Malley’s lab demonstrated that other molecules, completely unrelated to steroid hormones, could activate progesterone receptors in the rat brain (7). Apparently, the only reason why nobody before thought of an alternative mechanism of steroid hormone receptor activation was that the progesterone receptor has been discovered as a progesterone-binding protein, named after progesterone, and was always thought of as the receptor for progesterone. Although the methodology for the identification of alternative mechanisms of receptor activation was available for quite sometime, the mental inertia of the name prevented other researchers from making this key discovery.
Supplemental material Box 4: Example of mental inertia caused by improper controls.
Using incorrect standard controls is a danger for areas of biochemistry such as signal transduction, where any given molecule is networked in several different pathways and affected by several different stimuli. Thus, the use of an inadequate control stimulus will lead to an underestimation of this molecule’s role in other responses. For example, if one believes that a group of related toxins acts through the binding and activation of the same intracellular receptor, he/she still does not know how the different toxins bind the receptor in different ways, leading to different effects (8). Next, when studying a new chemical, he/she might ask whether the activation of the aforementioned receptor is part of its mechanism of action. This question might arise if this chemical is shown to bind the receptor. The use of the known control stimulus could lead to the belief that the mechanism of action of this compound is not mediated by the receptor in question, simply because the new compound and the standard control activate the receptor differently.
Supplemental material Box 5: When was apoptosis first observed?
The issue of cells being killed versus cells committing suicide came to light in 1972, when Kerr et al. described a specific series of cellular events, consistent with the idea of cells committing suicide, which they termed “apoptosis” (9). Since then, almost two hundred thousand papers have been published on the subject, and the concept of cellular suicide has become a staple of cell physiology.
Cellular death by apoptosis is highly organized – the cell kills itself and packages its fragments into tidy membrane-enclosed parcels that can be picked up and utilized by other cells. This formation of parcels from the cell body, which can be easily seen under the microscope, was what had caught the attention of and was described by John Kerr, Andrew Wyllie and Alastair Currie. Interestingly, images of cells undergoing apoptosis were observed and published multiple times, long before the 1972 publication by John Kerr and his colleagues. Yet the contemporary framework of thinking had prevented the authors, as well as the readers, from asking the right questions or coming to the right conclusions.
Supplemental material Box 6: Can microbes be transmitted through water?
In the IXX century, cholera epidemics caused thousands of deaths, and the mechanism of cholera transmission was investigated by many physicians. There were multiple competing explanations for the origins of this disease – dangerous elevation of the village relative to the sea level, water transmission, infection through inhalation of contaminated vapors (miasma theory), and even the germ theory--but the proponents of the germ theory were sure that germ transmission should occur through inhaling the contaminated air (10). Snow was the first to separate the idea that cholera is caused by germs alone from the idea that all infections occur through breathing in contaminated air (11). Unlike his colleagues, who either accepted or rejected these theories altogether, he advocated the theory of germs as the causative agents of cholera, but suggested (and brilliantly demonstrated) that the transmission occurs through contaminated water (10).
Supplemental material Box 7: Life saving technical mistake.
Airborne contamination was not at all uncommon in microbiology labs in the 1920s (in fact, it still happens today). If it happens too often, one’s experimental skills will be questioned. Usually, the contaminated plate is rapidly discarded and the experiment is repeated. But when this happened to Alexander Fleming in 1928, he took a closer look at the contaminated plate and noticed that mold was inhibiting bacterial growth. Through his further investigation of the mechanism of mold’s inhibitor effects on bacterial growth, he discovered penicillin, thus starting the era of antibiotics (12).
Supplemental material Box 8: Valid observation or a technical mistake: Nonyl-phenol mimics effect of steroid hormones.
A great example of scientists overcoming this type of metal inertia is the discovery of Anna Soto and Carlos Sonnenschein (13), who were studying the effects of estrogens on hormone-sensitive mammalian cells. Their experiments were conducted by placing cells in multiple wells of a tissue culture plate; some wells were exposed to varying amounts of hormones, others with vehicle controls only. Surprisingly, the untreated control cells sometimes looked exactly like the treated cells. Instead of discarding these results as a mix-up during hormone addition (which is not uncommon in experiments that involve multi-well plates), the scientists investigated this phenomenon. They found that the plastic plates, where the strange effect was taking place, release nonyl-phenol. This and similar compounds were routinely added to plastics to make them last longer and possess better qualities. Further investigation resulted in the discovery that some plastics release multiple chemicals that mimic the effects of sex hormones or "endocrine disruptors". This finding caused the re-evaluation of environmental safety standards and opened up a new research field on the potential effects of hand-made materials on human reproduction and hormone-dependent cancers (14,15).
Supplemental material Box 9: Did the Nobel Prize Committee overlook the difference between bacteria and immortal particles?
In their famous experiment, Salvador Luria and Max Delbruck demonstrated the random nature of mutations, laying the groundwork for the field of bacterial genetics which earned the scientists a Nobel Prize. The experiment dealt with the following phenomenon: E. coli are usually killed by a bacterial virus, bacteriophage T1. However, a small number of bacteria are resistant to being killed by this phage and form colonies on phage-containing Petri dishes. The researchers therefore asked if the resistant bacteria exist in the culture before exposure to the phage or if resistance is induced by the phage?
To answer this question, they used two groups of Petri dishes. Each dish was inoculated with the same number of bacterial cells exposed to the virus. However, all the bacteria in the first group of dishes were progeny of a single bacterial cell. By contrast, each dish in the second group received bacteria from one of twenty different sources. Thus, dishes in the first group were somewhat identical, whereas plates in the second group were genetically heterologous.
Luria and Delbruck assumed that if mutations determining resistance to the phage occur as a result of bacterial exposure to the phage, dishes in both groups would have the same number of resistant colonies (as the number of induced resistant colonies per plate would be proportionate only to the number of bacterial cells, which is equal for each plate in both groups). Contrarily, if the bacterial cells resistant to phage appear as a result of mutations preceding exposure to the virus, then each plate in the first group would have roughly the same number of bacterial colonies. At the same time, plates within the second group would differ significantly in the number of phage-resistant colonies. Such a finding would demonstrate that the frequency of phage-resistant mutations is a trait of each bacterial strain. The experiment resulted in a distribution of mutation frequencies that was much greater for the second group than the first group, thus supporting the idea that phage resistance is caused by a spontaneous mutation that occurs before the exposure to the phage. Interestingly, the experimental approach of Luria and Delbruck is somewhat similar to the one used in physics, where bacteria are viewed as “immortal objects”, like particles of an ideal gas.
Whereas their experiment demonstrated that resistance mutations arise spontaneously prior to T1 exposure, it did not answer the second part of the question: “Could the phage induce a resistance in bacteria?” If the frequency of this induction is much lower than that of spontaneous mutations, this experimental design would have not have revealed it. Moreover, a cell has a fundamental difference from Newtonian particles: cells can die. Exposure to the phage was lethal to bacteria. Thus, in all likelihood, bacteria would not have enough time to develop an adaptive mutation. This experimental approach was extrapolated to responses of other types of stress, and as a result, the discovery of adaptive evolution and stress-induced adaptive mutagenesis was postponed for decades (16).
Supplemental material Box 10: A great discovery, which was “hot” at the time.
A good example of a scientific problem that was “not hot enough” for its time, but eventually become a hot topic, is the 1956 work of Bruce Glick, Timothy S. Chang, and R. George Jaap (17), which has since revolutionized the field of immunology. In this classical work, Glick et al. showed that B-lymphocytes develop and mature in the bursa of Fabricius, a specialized immune organ of birds. This discovery provided a model system for the study of B-lymphocyte maturation and led to a better understanding of its role in the immune response. Interestingly, humans and other mammals lack this specialized organ, thus making it impossible to study this process at that time.
Glick tried to publish his discovery in Science and other reputable journals, but to no avail (18). Probably, reviewers at the time considered this seminal work to be of little scientific significance, because nobody was interested in studying the immunology of birds as a model for medicine. In the end, he published his work in a journal that was neither prestigious nor medical – Poultry Science—a publication which dealt mostly with matters of the poultry industry and was not available at most medical school libraries. However, this little paper probably had a larger impact on the development of medicine than many volumes of prestigious medical journals combined.
Supplemental material Box 11: Shifting the paradigm.
The Universal Genome hypothesis proposed by Michael Sherman (19) suggests a historical paradigm shift in our current views of evolution if it were to be experimentally tested and proven. For a long time, the underlying logic of evolutionary thinking was ‘new useful functions appeared after organisms acquire or develop genes necessary to perform them’. Thus, the major path to understanding evolution was to elucidate mechanisms of emergence of new genes and gene forms appearing when and where they are necessary. This line of thinking can form a good working hypothesis for the most simple events, but it does not work for complex systems like signal transduction cascades. Indeed, the appearance of all but one element of a signaling pathway would be as dysfunctional as a watch without just one cogwheel. Much effort was devoted to explaining the gradual appearance of all the elements of signal transduction pathways one-by-one prior to their assembly into working mechanisms. Contrary to this view, however, Sherman hypothesized that all the necessary genes and signaling pathways are silently present in the early organisms, where they perform no function. This notion shifts the major question to ‘what makes these silent and useless pathways become active and crucial elements of the cell’s and organism’s physiology?’ Notably, Sherman’s Universal Genome hypothesis does not resolve the bottlenecks of the preexisting paradigm; rather, it just does not face them.