The Bac­te­r­ial Chro­mo­some: A Phys­i­cal Biologist's Apol­ogy

A Per­spec­tive by Suck­joon Jun

I en­tered the bac­te­r­ial chro­mo­some field in 2004 as a fresh Ph.D. trained in the­o­ret­i­cal physics. Ten years is not long enough for one to gain the depth and breadth of a sci­en­tific dis­ci­pline of long his­tory, cer­tainly not for an early ca­reer sci­en­tist to write an es­say of the sta­tus of A Mathematician's Apol­ogy (Hardy 1940). Nev­er­the­less, I agreed to write this Per­spec­tive as a physi­cist who en­tered bi­ol­ogy, be­cause my col­leagues are of­ten cu­ri­ous to know what dri­ves physi­cists to be­come (phys­i­cal) bi­ol­o­gists, and make them stay in bi­ol­ogy de­spite many chal­lenges. I also wanted to share sev­eral lessons I have learned be­cause, while some of them are per­sonal and spe­cific to my field, I have a good rea­son to be­lieve that they might res­onate with many fu­ture trav­el­ers. This Per­spec­tive is for them.

I would like to start with the story of one of the most fa­mil­iar and yet mys­te­ri­ous forces in na­ture—grav­ity. Galileo is said to have dropped two balls of dif­fer­ent masses from lean­ing Tower of Pisa in Italy some five hun­dred years ago. His ex­per­i­ment was to demon­strate that, on the con­trary to Aristotle's the­ory, the falling rate of the balls was in­de­pen­dent of their mass. A mod­ern ver­sion of this ex­per­i­ment was per­formed on the Moon by the Com­man­der of Apollo 15 with a ham­mer and a feather. For a movie of this ex­pe­ri­ence, click here. When re­leased from the same height at the same time, the two falling bod­ies hit the sur­face of the Moon si­mul­ta­ne­ously! On the Earth, how­ever, the feather would have flut­tered, as if alive, be­cause of the air.

Ini­tially at­tracted to the beauty of Am­s­ter­dam, I started my post-doc­toral re­search at AMOLF, an in­ter­dis­ci­pli­nary re­search in­sti­tute known for ex­cit­ing in­ter­ac­tions at the in­ter­face be­tween phys­i­cal and bi­o­log­i­cal sci­ences. The forces I was in­ter­ested in were much less tan­gi­ble than grav­ity. In par­tic­u­lar, I was sup­posed to ex­plain the dri­ving force un­der­ly­ing seg­re­ga­tion of a repli­cat­ing chro­mo­some in Es­cherichia coli. It sounded sim­ple to me, ex­cept that I barely knew any­thing about bac­te­ria, cer­tainly with­out re­al­iz­ing that it was one of the long-stand­ing prob­lems in bi­ol­ogy. I knew the DNA bio­physics lit­er­a­ture fairly well, but when I saw the beau­ti­ful 1992 il­lus­tra­tion of E. coli in Good­sell, it was ob­vi­ous that some­thing like the worm­like chain model was not go­ing to be very use­ful to un­der­stand seg­re­ga­tion of the whole chro­mo­some. What wor­ried me was the directionality—if I were a small pro­tein sit­ting on a repli­cat­ing chro­mo­some, could I tell which DNA seg­ment be­longs to which sis­ter DNA? Physi­cists like ques­tions like that, whether they are rooted in physics or bi­ol­ogy.

The first chal­lenge I faced was some­thing un­ex­pected. It is the as­pect of "nat­ural his­tory" in bi­ol­ogy. In physics, pa­pers are rel­a­tively straight­for­ward to judge their va­lid­ity and im­por­tance. Physi­cists find the bi­ol­ogy lit­er­a­ture much less so, and they can be taken aback when they learn that the pub­lished data they mod­eled are later proven wrong. It can take a long time for physi­cists to learn which bi­ol­o­gists are more re­li­able than oth­ers and who is more care­ful about what they choose to say about their data. This nat­ural his­tory as­pect of bi­o­log­i­cal re­search is some­thing one can­not learn from the lit­er­a­ture, as it is more about ex­pe­ri­ence than in­tel­li­gence. If the physi­cist learns bi­ol­ogy from an­other physi­cist, there is a dan­ger of learn­ing a lan­guage and cul­ture from a non-na­tive speaker. Al­ter­na­tively, if he goes to a bi­ol­ogy con­fer­ence to learn (as­sum­ing he al­ready knows which con­fer­ence to go), typ­i­cally the bi­ol­ogy is over­whelm­ing even for an ex­pe­ri­enced bio­physi­cist. The best so­lu­tion is to find re­li­able and gen­er­ous bi­ol­o­gists col­leagues, who are not only will­ing to walk the physi­cist through the bi­ol­ogy in an hon­est man­ner, but also gen­uinely in­ter­ested in think­ing about the prob­lem from a dif­fer­ent an­gle with the physi­cist. In this re­gard, I have been ex­tremely for­tu­nate to have worked with many gen­er­ous bi­ol­o­gists in the field, in par­tic­u­lar, Stu­art Austin, Con­rad Woldringh, and An­drew Wright.

Fig­ure 1. The spin­dle para­dox in E. coli. At the slow­est growth rates, the chro­mo­some rests in B/G1 be­fore the repli­ca­tion pe­riod C/S starts. This is as in eu­kary­otes, and in prin­ci­ple hy­po­thet­i­cal spin­dles can sep­a­rate the du­pli­cat­ing ori re­gions (1 and 1') to the op­po­site poles of the E. coli cell (red ar­rows). At faster growth rates, how­ever, repli­ca­tion is con­tin­u­ous and mul­ti­ple rounds of repli­ca­tion cy­cles over­lap. The il­lus­tra­tion on the right shows four copies of du­pli­cates Ori's. The hy­po­thet­i­cal spin­dles in this case will make the ori 1' and ori 2 swap their po­si­tions. That is, the spin­dles would even­tu­ally mix the chro­mo­somes, rather than sep­a­rate them (adopted from Youn­gren et al. 2014)

The im­por­tant bi­ol­ogy I learned early on from Con­rad Woldringh re­garded the phys­i­ol­ogy and cell cy­cle of E. coli. Con­rad was par­tic­u­larly fond of stud­ies in the 1950s and 1960s, such as the ones from the Copen­hagen School of Bac­te­r­ial Phys­i­ol­ogy (Maaløe and Kjeldgaard, 1966). Con­sider, for in­stance, E. coli grow­ing in nu­tri­ent rich medium, whose gen­er­a­tion time is shorter than the du­ra­tion of DNA repli­ca­tion. Since there must be a strict one-to-one cor­re­spon­dence be­tween repli­ca­tion ini­ti­a­tion and cell di­vi­sion (Mitchi­son 1971), a new round of cell cy­cle must start be­fore the pre­vi­ous round of cell cy­cle is com­pleted. Fig­ure 1 il­lus­trates a schematic de­vel­op­ment of over­lap­ping cell cy­cles in E. coli. Why is this im­por­tant? For one thing, we can en­ter­tain a gedanken­ex­per­i­ment about the role of hy­po­thet­i­cal mi­totic spin­dle in bac­te­ria as shown in Fig­ure 1. Un­der mul­ti­fork repli­ca­tion, the hy­po­thet­i­cal spin­dles that would pull two pairs of cen­tromeres would mix the sis­ter chro­mo­somes, rather than sep­a­rate them! From this "spin­dle para­dox," I learned my first sci­en­tific les­son in the bac­te­r­ial chro­mo­some. That is, to re­ally un­der­stand chro­mo­some seg­re­ga­tion, one should pay at­ten­tion to the global, phys­i­cal prop­er­ties of the chro­mo­some it­self.

The poly­mer model of the bac­te­r­ial chro­mo­some Bela Mul­der and I pro­posed in 2006  (Jun and Mul­der, 2006) was based on the first les­son de­scribed above. Our two main ideas were that (1) when the right phys­i­cal con­di­tions are met, poly­mers with ex­cluded-vol­ume in­ter­ac­tions will spon­ta­neously un­mix with each other even in strongly con­fined space such as E. coli's cel­lu­lar vol­ume (2) DNA will move much faster and more freely in the cy­to­plas­mic space than in­side the nu­cleoid, i.e., the chro­mo­some dy­nam­ics should be con­sid­ered in 3D, not in 1D. Some of the data we had mod­eled soon had to be up­dated by new data (Wang et al. 2006; Nielsen et al. 2006), but these two ideas were ro­bust to the spe­cific fea­tures of the data. How­ever, an im­por­tant ques­tion re­mained: what are the right phys­i­cal con­di­tions for spon­ta­neous seg­re­ga­tion? It took sev­eral years to an­swer the ques­tion in a se­ries of the­ory pa­pers (sum­ma­rized in Jun and Wright, 2010).

The next im­por­tant les­son was the re­sult of an iden­tity cri­sis. At one point af­ter hav­ing worked out the most of the ini­tial chro­mo­some mod­el­ing, I saw real E. coli for the first time un­der the mi­cro­scope in Conrad's lab across the street. They were very tiny, and they were so alive. Watch­ing the tiny cells swim­ming and tum­bling with my own eyes was a hum­bling ex­pe­ri­ence; it made me re­al­ize I had no real un­der­stand­ing of what is meant by liv­ing or­gan­isms. Even­tu­ally, I un­der­stood the fun­da­men­tal dif­fer­ence be­tween bi­o­log­i­cal physics and bi­ol­ogy, that bi­ol­ogy in its heart is an ex­per­i­men­tal sci­ence and that, in bi­ol­ogy, de­tails—"pref­ac­tors" as op­posed to "ex­po­nents"—re­ally mat­ter. I be­lieve vir­tu­ally all the­o­ret­i­cal bio­physi­cists go through a pe­riod of a sim­i­lar cri­sis, and ask them­selves how deep they should go into the bi­ol­ogy. In my case, I came to a con­clu­sion that I re­ally wanted to un­der­stand E. coli, and it was only with a great for­tune that I was given the free­dom and re­sources to pur­sue my ex­per­i­men­tal ideas while a Bauer Fel­low at Har­vard (Pel­letier et al. 2012). For any physi­cist read­ers want­ing to un­der­stand the bi­o­log­i­cal re­al­ity and why their bi­ol­o­gist col­leagues think the way they do, I highly rec­om­mend to ask them to ad­mit to their lab, at least for one sum­mer. The ex­pe­ri­ence will change their view per­ma­nently.

Fig­ure 2. Il­lus­tra­tion of the E. coli chro­mo­somes dur­ing mul­ti­fork repli­ca­tions based on data in Youn­gren et al. 2014. By David Good­sell.

I de­scribed at the be­gin­ning why the gen­eral mech­a­nism of chro­mo­some seg­re­ga­tion and or­ga­ni­za­tion should ac­count for all growth and cell cy­cle con­di­tions, and that it must come from the chro­mo­some it­self, not from the spin­dles. The fi­nal and most re­cent les­son is in­deed about the E. coli chro­mo­some dur­ing its gen­eral, over­lap­ping cell cy­cles in liv­ing cells. This is due to the heroic ef­fort of Stu­art Austin (Youn­gren et al. 2014). In 2006, two lab­o­ra­to­ries led by Stu­art Austin and David Sher­ratt pub­lished im­por­tant re­sults on the or­ga­ni­za­tion of the E. coli chro­mo­some un­der slow­est-grow­ing con­di­tions, un­der which the cell cy­cles do not over­lap (Wanget al. 2006; Nielsen et al. 2006). The two pa­pers re­ported that, un­like the pre­vi­ous be­lief in the field, the left (L) and right ® arms of the E. coli chro­mo­some oc­cupy each cell halves along the long-axis of the cell, such that the repli­cated chro­mo­somes are or­ga­nized as LRLR or LRRL. Stu­art, how­ever, did not stop there. His group started tack­ling what had been con­sid­ered im­pos­si­ble. They started the mea­sure­ments and analy­sis of in­tra­cel­lu­lar po­si­tions of cou­ple dozens of dual ge­nomic loci mark­ers un­der over­lap­ping cell-cy­cle con­di­tions. This is a daunt­ing task be­cause, for any given mo­ment, every cell con­tains sev­eral ho­mol­o­gous copies of each ge­nomic lo­cus, and de­ci­pher­ing the or­ga­ni­za­tion and dy­nam­ics of the whole chro­mo­some based on their po­si­tional in­for­ma­tion was some­thing that had never been done be­fore. Nev­er­the­less, Stuart's group re­lent­lessly pushed their ef­forts with­out pub­lish­ing any­thing for sev­eral years. When the task was done, the end re­sult was sim­ple, el­e­gant, and sur­pris­ing. The or­ga­ni­za­tion of the chro­mo­some dur­ing mul­ti­fork repli­ca­tion was that of a sim­ple branched donut, with the two arms of the chro­mo­some oc­cu­py­ing each cell halves along the ra­dial axis of the cell. This re­sult could not have been pre­dicted based on our knowl­edge of slow­est-grow­ing cells, but it en­com­passes all the pre­vi­ously known re­sults and move be­yond them into some­thing new and gen­eral (Youn­gren et al. 2014). David Good­sell has kindly up­dated his E. coli il­lus­tra­tion based on the data. (Fig­ure 2).

So what then does grav­ity have to do with the bac­te­r­ial chro­mo­some? I think it's re­mark­able that we ac­cept the ex­is­tence of an in­vis­i­ble, at­trac­tive force from the Earth, even when a sheet of pa­per flut­ters or even flies away in the wind. Per­haps it's the tan­gi­bil­ity of grav­ity that helps us feel the con­cept. To me, phys­i­cal forces in­trin­sic to the poly­meric na­ture of the chro­mo­some are as mys­te­ri­ous and yet tan­gi­ble as grav­ity. Mol­e­cules, be they "air mol­e­cules" or pro­teins, are im­por­tant in un­der­stand­ing the de­tails of the mo­tions. How­ever, un­der­stand­ing the ma­jor phys­i­cal dri­ving forces, and how physics and bi­ol­ogy have worked to­gether dur­ing the course of evo­lu­tion, will only deepen our ap­pre­ci­a­tion of the role of the mol­e­cules. Re­cent work by Wiggins's group shows an ex­em­plary ap­proach to­wards such a di­rec­tion (Kuwada et al. 2013). As for me, I am con­vinced that the (in)tangible phys­i­cal forces drive chro­mo­some seg­re­ga­tion no mat­ter what, and physi­cists will con­tinue to en­ter bi­ol­ogy un­til one day the bound­ary be­tween the two dis­ci­plines dis­ap­pears.

Book Ref­er­ences

Good­sell DS (1992) The Ma­chin­ery of Life, Springer, Berlin

Hardy GH (1940) A Mathematician's Apol­ogy, Cam­bridge Uni­ver­sity Press, Cam­bridge

Maaloe O, Kjeldgaard NE (1966) Con­trol of macro­mol­e­c­u­lar syn­the­sis: A study of DNA, RNA, and pro­tein syn­the­sis in bac­te­ria, W. A. Ben­jamin

Mitchi­son JM (1971) The Bi­ol­ogy of Cell Cy­cle, Cam­bridge Uni­ver­sity Press, Cam­bridge

Suck­joon is an As­sis­tant Pro­fes­sor at the De­part­ment of Physics and Sec­tion of Mol­e­c­u­lar Bi­ol­ogy in Di­vi­sion of Bi­ol­ogy, Uni­ver­sity of Cal­i­for­nia San Diego.