Mutations (changes in DNA): the lifeblood of genetic analysis hence most mutations with any functional consequence (mutant phenotype) disrupt normal (wildtype) gene function fact: it’s easier to mess things up than to make them better

Mutations (changes in DNA): the lifeblood of genetic analysis Geneticists make mutations that disrupt normal gene functions (thereby creating a functional difference between alleles) to discover what genes do what (…and sometimes geneticists can learn a thing or two about “how” even before biochemists enter the picture) …so they can tell biochemists where to look to learn how those genes do what they do

Mutations (changes in DNA): the lifeblood of genetic analysis Not all mutant phenotypes are equally informative.

Fig wildtype Krüppel hunchbackknirps examples of some maximally informative mutant phenotypes for understanding metazoan pattern formation

Forward genetics: Reverse genetics: ultimately determine wildtype molecules ultimately determine mutant phenotype (functional consequences of disruption) to infer wildtype function start with mutant phenotype (functional consequences of disruption) to infer wildtype gene function start with wildtype molecules (ultimately transcription units a.k.a. genes) (learn molecular mechanism) (…and learn molecular mechanism)

Mutations: the lifeblood of genetic analysis …to infer wildtype gene functions Not Mendel’s goal: Not Morgan’s goal: (predict the appearance and breeding behavior of hybrids) (learn how inherited information is transmitted & how it changes) (neither had any hope of discovering what gene are)

“What are the genes? What is the nature of the elements of heredity that Mendel postulated as purely theoretical units? … Frankly, these are questions with which the working geneticist has not much concern himself… T.H. Morgan The Relation of Genetics to Physiology and Medicine Nobel Lecture, June 4, 1934

Mutations: the lifeblood of genetic analysis Herman Muller (1930s): inferred how mutations can affect gene functioning. Muller: spontaneous & radiation-induced Muller: exploited giant polytene chromosomes & invented balancer chromosomes (functional categories) (mutagenesis) (mutant screens & selections) (Nobel Prize 1946) a Morgan "student": (1) What kinds can we make? (2) How do we make them? (3) How do we find them?

Muller categorized mutations with respect to change in gene function relative to wildtype Loss-of-(wildtype)function (l-o-f) mutant alleles “Gain”-of-(over wildtype)function (g-o-f) mutant alleles generally recessive (a + /a - : one functional copy “suffices”) clear exception: l-o-f mutations in haploinsufficient genes are dominant by definition (Minute mutations in flies; Df(M)/+=M/+) (2) so long as we don’t look too hard (1) so long as all other genes ok) lof alleles causing cancer: identify tumor suppressor genes generally dominant Antp X /Antp + gof alleles causing cancer: identify proto-oncogenes (often misexpression, wrong time/place): fly leg where antenna should be

Loss-of-(wildtype)function (l-o-f) mutant alleles “Gain”-of-(over wildtype)function (g-o-f) mutant alleles partial lof: hypomorph(ic) (leaky) Important goal of genetic analysis: define the null phenotype phenotypic series: (helpful with pleiotropy) set of alleles with progressively less function complete lof: amorph(ic) (null) too much of a good thing: hypermorph(ic) something new & different: neomorph(ic): different in kind (e.g. fusion protein) ectopic expression (dominant negative) wrong time/place: antagonizes (poisons)wildtype: antimorph(ic)

Loss-of-(wildtype)function (l-o-f) mutant alleles “Gain”-of-(over wildtype)function (g-o-f) mutant alleles partial lof: hypomorph(ic) (leaky) complete lof: amorph(ic) (null) too much of a good thing: hypermorph(ic) something new & different: neomorph(ic): antagonizes (poisons)wildtype: antimorph(ic) Muller did not just define the five basic ways the functioning of a gene can be changed by mutation (without, he noted, changing its ability to faithfully replicate) He gave us operational tests to determine to which class a given mutant allele might belong

increased dose of mutant, phenotype more wildtype Muller’s tests: how does the phenotype change when you: (1) hold the number of mutant alleles constant and change the number of wildtype alleles. (2) hold the number of wildtype alleles constant and change the number of mutant alleles. w a /w a (darker, more “wildtype”) > w a /Df(w) w a /Y (lighter, less “wildtype”) < w a /Y; Dp(w a )/+ 2 copies w a allele 1 copy w a allele 2 copies w a allele the white gene started it all, and has kept it all going to this day w apricot w+w+ w-w-

increased dose of wildtype, phenotype more wildtypeincreased dose of mutant, phenotype more wildtype See how the phenotype changes when you: (1) hold the number of mutant alleles constant and change the number of wildtype alleles. (2) hold the number of wildtype alleles constant and change the number of mutant alleles. w a /w + (darker, more “wildtype”) > w a /Df(w) w a /Y (lighter, less “wildtype”) < w a /Y; Dp(w + )/+ 1 copy w + allele 0 copies w + allele 1 copy w + allele

w apricot Loss-of-(wildtype)function (l-o-f) mutant alleles “Gain”-of-(over wildtype)function (g-o-f) mutant alleles partial lof: hypomorph(ic) (leaky) complete lof: amorph(ic) (null) too much of a good thing: hypermorph(ic) something new & different: neomorph(ic): antagonizes (poisons)wildtype: antimorph(ic)

partial lof: hypomorph(ic) Increased dose of mutant: phenotype more wildtype Increased dose of wildtype: phenotype more wildtype w apricot Muller thought so, and realized that he had discovered X-chromosome dosage compensation: XX=XY (more about that later) w a /w a = (same color as) w a /Y But isn’t in RATHER curious then that: O + O / v

Increased dose of mutant: No change in phenotype Increased dose of wildtype: phenotype more wildtype partial lof: hypomorph(ic) Increased dose of mutant: phenotype more wildtype Increased dose of wildtype: phenotype more wildtype w apricot complete lof: amorph(ic) w - The characterization of a mutant allele as a amorphic (null) vs. hypomorphic is generally made based only on a comparison of the homozygous mutant to the hemizygous mutant (and with the knowledge that the mutant is recessive): white X /white X vs. white X /Df(w) Truth be known, geneticists take shortcuts (c.f. cis/trans test) …potential pitfalls, but a good place to start

too much of a good thing: hypermorph(ic) something new & different: neomorph(ic): antagonizes (poisons)wildtype: antimorph(ic) Increased dose of mutant: No change or more mutant in phenotype Increased dose of wildtype: No change in phenotype Increased dose of mutant: phenotype more mutant Increased dose of wildtype: phenotype more mutant Increased dose of mutant: phenotype more mutant (if possible) Increased dose of wildtype: phenotype more wildtype “Gain”-of-(over wildtype)function (g-o-f) mutant alleles our friend from the X-files, Antp ovo D1

to infer the normal function of a gene: LOF alleles GOF alleles simplest to interpret usually more interesting but generally harder to interpret.. especially if don’t know amorph (null) phene. ovo D(ominant)#1 dominant female-specific sterile antimorph ovo e8K recessive female-specific sterile hypomorph How can we determine whether ovo D1 and ovo e8K are alleles? consider the fly gene ovo as a good example of the problems that arise: complementation test? (ovo D1 /ovo e8K ?) Both X-linked & similar phenotypes, but can't directly map ovo D1

Can use GOF alleles to generate LOF alleles relatively easily: …then establish allelism using the GOF-LOF double mutant alleles -- also helps define the null phenotype “revert” (suppress) the dominance of a GOF allele LOF allele How can we determine whether ovo D1 and ovo e8K are alleles? (make it stop doing something bad by inducing LOF mutation IN CIS)

ovo D1 / + tumorous female germline (dominant) can't use the complementation test: ovo D1 / ovo e8K wildtype or mutant? neither would tell us anything not haplo- insufficient GOF ovo e8K / + ovo e8K / ovo e8K tumorous female germline (recessive) wildtype female germline LOF Df(ovo) / + wildtype female germline { }

cis-trans test? …if in same gene, cis phenotype ≠ trans …if in different genes, cis phenotype = trans ovo D1 + / + ovo e8K ovo D1 ovo e8K / + + tumorous less tumorous or wt. ovo D1 + / + ovo e8K ovo D1 ovo e8K / + + tumorous

…if in same gene, cis phenotype ≠ trans ovo D1 + / + ovo e8K ovo D1 ovo e8K / + + tumorous less tumorous or wildtype ovo D1 + / + ovo -null ovo D1 ovo -null / + + tumorous wildtype but how construct? Consider a slightly different cis/trans situation : ovo D1 + / + + tumorous ( )

ovo D1 + / + ovo -null ovo D1 ovo -null / + + tumorous wildtype mutagenize to “revert” the dominance but what if true reversion : + +/+ + ? Consider a slightly different cis/trans situation : double mutant homozygote will be tumorous ovo D1 + / + + tumorous ovo D1 ovo -null / + + wildtype what we have: what we want to generate: look at homozygote ovo D1 ovo -null /

mutagenize to “revert” the dominance now do complementation test to distinguish the alternatives: ovo D1 ovo -null / + ovo e8K + + / + ovo e8K vs. tumorous wildtype ovo D1 + / + + tumorous ovo D1 ovo -null / + + wildtype what we have: what we want to generate:

mutagenize to “revert” the dominance ovo D1 ovo -null / + ovo e8K + + / + ovo e8K vs. tumorous wildtype ovo D1 + / + + tumorous ovo D1 ovo -null / + + wildtype..hence we have established that ovo D1 and ovo e8K are functional alleles Failure to complement with ovo e8K means that we have to lose the gene function that ovo e8K is missing in order to “revert” (suppress) the dominance of ovo D1. …allowed us to do a complementation test:

Same for Antp 32a5 and Ns 1, two GOF mutant alleles that cause the antenna to develop as a leg instead. (I owe my own success to reverting dominant mutations) revert their dominance (one gets a recessive lethal in each case) …the resulting recessive lethal mutant chromosomes fail to complement.