Of Terms in Bi­ol­ogy: Moon­light­ing

by Christoph

The Cam­bridge Dic­tio­nary de­fines "moon­light­ing" as:

the act of work­ing at an ex­tra job, es­pe­cially with­out telling your main em­ployer

(Far) off the job mar­ket, "moon­light­ing pro­teins" are sin­gle polypep­tide chains that per­form mul­ti­ple bio­chem­i­cal or bio­phys­i­cal func­tions in ar­chaeal, bac­te­r­ial, and eu­kary­otic cells. Con­stance Jef­fery coined this term in 2003 to ac­count for the re­cur­rent ob­ser­va­tion that newly iden­ti­fied pro­teins in­volved in the com­plex processes of repli­ca­tion, tran­scrip­tion, and tu­mor sup­pres­sion (in euks) had other, pre­vi­ously iden­ti­fied, func­tions, of­ten in cen­tral me­tab­o­lism, for ex­am­ple in gly­co­lysis or the TCA cy­cle. But who is, in this case, the 'main em­ployer' from the Cam­bridge Dic­tio­nary de­f­i­n­i­tion above? Must be the sci­en­tist, sur­prised to find a novel func­tion for a well-known pro­tein in an un­sus­pected con­text, right? This is a bi­ased per­cep­tion that started with George W. Bea­dle & Ed­ward L. Tatum's "One gene – one en­zyme hy­poth­e­sis" (based on their sem­i­nal pa­per from 1941), and be­fore it was clear that DNA is the ge­netic ma­te­r­ial (Av­ery et al. was pub­lished in 1944). "Bi­as­ed per­cep­tion" be­cause the hy­poth­e­sis tac­itly con­tained the as­sump­tion that a pro­tein, en­co­ded by a gene, would have only one func­tion, namely the one that in­ter­ested re­searchers at that mo­ment.

Fig­ure 1. Qβ repli­case con­sists of the virus-en­­coded RNA-de­pen­dent RNA poly­merase (β‑sub­unit, yel­low), and three host trans­la­tion fac­tors: EF-Tu (red), EF-Ts (blue), and ri­bo­so­mal pro­tein S1 (green). Sim­pli­fied schematic de­pic­tions of Qβ RNA (black) repli­ca­tion ini­tiation (left) and ter­mi­na­tion (right). The OB3 do­main of S1 is mo­bile and ca­pa­ble of piv­otal ro­ta­tion, us­ing the α he­lix con­nect­ing OB2 and OB3 as a swing arm. Source

An ex­am­ple from the early 1970s of pro­teins with dual func­tions comes from stud­ies on the mech­a­nism of re­plication of the small single‑stranded RNA bac­te­rio­phage Qβ, which re­cruits host pro­teins in­volved in trans­la­tion to repli­cate its RNA. Qβ RNA repli­ca­tion re­quires the phage-en­coded repli­case – an RNA-de­pen­dent RNA poly­merase – in com­plex with the host's ri­bo­so­mal pro­tein S1, and trans­la­tional elon­ga­tion fac­tors EF-Tu and EF-Ts (Fig­ure 1). More re­cent struc­tural stud­ies show that EF-Tu and EF-Ts do not un­dergo gross con­for­ma­tional changes in the Qβ repli­case com­plex, and con­tact the phage RNA much the way they con­tact tR­NAs dur­ing trans­la­tion. Al­so, ri­bo­so­mal pro­tein S1 "fits" in the com­plex in a way that cor­re­sponds to its role dur­ing trans­la­tional ini­ti­a­tion (S1 is im­por­tant for ini­ti­at­ing trans­la­tion in mR­NAs lack­ing a Shine-Dal­garno se­quence). Among phages, "recruit­ing" one or more host repli­ca­tion pro­teins for the replica­tion of their genomes is rather the rule than the excep­tion, but re­cruit­ing host trans­la­tion pro­teins for this pur­pose is at least un­usual. But with phages you al­ways have to be open for sur­prises, and in the 1970s, dual func­tions of pro­teins were thought of ex­cep­tions from the "one gene – one pro­tein (one func­tion)" hy­poth­e­sis rather than its fal­si­fi­ca­tion.

En­zymes can per­form func­tions other than the ones they are known for – "moon­light­ing" accord­ing to Jef­fery – when they are "ge­o­graph­i­cally dis­placed" from their reg­u­lar com­part­ment. For ex­ample, eno­lase (the en­zyme that car­ries out the ninth step of gly­col­y­sis) is of­ten found on the sur­face of path­o­genic or com­men­sal bac­te­ria where it in­ter­acts with the host's ex­tra­cel­lu­lar ma­trix or serves as a re­cep­tor for plas­mino­gen. Hen­der­son and Mar­tin (2011) give an overview of nor­mally cy­to­plas­mic bac­te­r­ial pro­teins that, when lo­cated on the cell sur­face, as­sume a role as vir­u­lence fac­tors. The lo­cal­iza­tion of cy­to­plas­mic pro­teins on the cell sur­face could eas­ily be a prepa­ra­tion or pu­rifi­ca­tion ar­ti­fact, for ex­am­ple bind­ing of pro­teins from dead and lysed cells to liv­ing sib­lings. How­ever, Saad et al. showed that, in the pres­ence of low glu­cose lev­els, L. plan­tarum has an ele­va­ted level of cell wall-as­so­ci­ated glyc­er­alde­hyde 3‑phosphate de­hy­dro­ge­nase (GAPDH, the en­zyme that nor­mally car­ries out the sixth step of gly­col­y­sis).

Fig­ure 2. A B. sub­tilis 168 ΔcpgA mu­tant is into­xi­cated by glu­cose or glu­conate. Disk dif­fu­sion as­says to mea­sure glu­cose and glu­conate sensiti­vi­ty were de­ter­mined on MH medium. Un­der our con­di­tions, 1% glu­cose (10 mg) is 55 μmols and 1% glu­conate is 46 μmols on the fil­ter. Source

In a very re­cent study, Sachla and Hel­man (2019) show that it is not nec­es­sary for a moon­light­ing pro­tein to be "ge­o­graph­i­cally dis­placed" to per­form dual func­tions. The cy­to­plas­mic B. subti­lis CpgA pro­tein be­longs to a sub­set of the so-called TRAFAC GT­Pases – of al­to­gether 17 P‑loop type GT­Pases in this bac­terium – that are in­volved in ri­bo­some as­sem­bly. CpgA is specif­i­cally in­volved in the as­sem­bly of the 30S sub­unit. A ΔcpgA mu­tant dis­plays a range of growth phe­no­types, in­clud­ing cold-sen­si­tiv­ity and in­creased sen­si­tiv­ity to­wards ri­bo­some-tar­geted an­tibiotcs (among other, not ri­bo­some-re­lated phe­no­types like al­tered cell mor­phol­ogy). Un­ex­pect­edly, this ΔcpgA mu­tant was un­able to grow in the pres­ence of glu­cose or glu­conate as mon­i­tored by a zone-of-in­hi­bi­tion as­say on Mueller-Hin­ton (MH) medium, a rich medium with abun­dant amino acids that sup­port glu­co­neogenic growth (Fig­ure 2). This un­usual phe­no­type – B. sub­tilis mu­tants that are in­tox­i­cated by glu­cose are not known – could be fully com­ple­mented by re­introduction of cpgA at its lo­cus, and also by over­pro­duc­tion of GndA (6‑phosphogluconate dehy­drogenase), its par­a­log GntZ, phos­pho­glu­cose iso­merase (PGI), and by yeast Pho13, an en­zyme that specif­i­cally ca­tab­o­lizes ery­throse-4-phos­phate (4PE). To un­der­stand the ra­tio­nale of these com­ple­men­ta­tions one has to dive deeper into gly­col­y­sis and the pen­tose-phos­phate path­way (PPP) (here is a meta­bolic flow chart). In brief, the ΔcpgA mu­tant is un­able to grow in the pres­ence of glu­cose or glu­conate be­cause a meta­bolic re­stric­tion arises from the in­hi­bi­tion of 6‑phospho­glu­conate de­hy­dro­ge­nase (GndA) by 4‑phosphoerythronate (4PE), an ad­ven­ti­tious ox­i­da­tion pro­duct that arises from the pen­tose phos­phate path­way (PPP) in­ter­me­di­ate ery­throse-4-phos­phate (4PE). 4PE ini­ti­ates an in­hi­bi­tion cas­cade in which 6‑phosphogluconate ac­cu­mu­lates and compe­ti­ti­ve­ly in­hibits phos­pho­glu­cose iso­merase (PGI), lead­ing to meta­bolic grid­lock, bac­te­riosta­sis, and ul­ti­mately cell death. CpgA as a broad-spec­trum phos­phatase pre­vents meta­bolic in­tox­i­ca­tion by cleans­ing the metabo­lite pool of po­ten­tially toxic mol­e­cules, in­clud­ing 4PE. The au­thors con­clude that the metabo­lite-proof­read­ing role of CpgA is es­sen­tial for growth of B. sub­tilis in the pres­ence of glu­cose or glu­conate, and that the role in ri­bo­some as­sem­bly seems im­por­tant for growth at low tem­per­a­tures where ri­bo­some as­sem­bly is po­ten­tially com­pro­mised.

A con­clu­sion? The term "moon­light­ing" may be a mis­nomer be­cause it is of­ten not pos­si­ble to dis­criminate in a mean­ing­ful way be­tween ma­jor and ac­ces­sory roles of a pro­tein in the dif­fer­ent path­ways it is in­volved in. But the term will stick, and rightly so. It will serve as a re­minder for sys­tems bi­ol­o­gists that net­work re­con­struc­tions based on, for ex­am­ple, pro­teome data have to ac­count for pro­teins 'work­ing at an ex­tra job, with­out telling their main em­ployer' in ad­di­tion to their 'an­no­tated' job (in the pro­tein data bases).

 

Other Posts

  • Push­ing the Ther­mo­dy­namic En­ve­lope into the Pro­teomic Edge

    by Tracey Mc­Dole — The word mar­ginal means to be at the outer or lower lim­its; min­i­mal for re­quire­ments; al­most in­suf­fi­cient. Cer­tainly be­ing mar­ginal sounds bad, but is it al­ways? In a re­cent is­sue of PNAS, Dill and col­leagues show that the pro­teome – a cell's col­lec­tion of thou­sands of dif­fer­ent types of pro­teins – is…

  • Hedg­ing Your Bets

    by Merry Youle — Bac­te­ria that are born ge­net­i­cally equal aren't nec­es­sar­ily the same. The same genome, re­sid­ing in cells side-by-side in the same medium in the same flask, does not guar­an­tee the same phe­no­type. One ex­am­ple that comes to mind is the per­sis­ters in E. coli pop­u­la­tions − the small num­ber of cells that…

  • E. coli Keeps Its Pow­der Dry

    by S. Mar­vin Fried­man — Whether in­hab­it­ing soil, fresh wa­ter, or ma­rine ecosys­tems, bac­te­ria are con­stantly fac­ing the threat of nu­mer­ous and ef­fec­tive preda­tors such as pro­tists, ne­ma­todes, or phages. To de­fend against such pre­da­tion, bac­te­ria have evolved a num­ber of strate­gies, in­clud­ing get­ting larger in size, mov­ing faster, pro­duc­ing de­fen­sive sec­ondary metabo­lites, and form­ing…

  • Phages Un­leashed: The Heist of CRISPR Re­peats

    by Sarah Ca­mara-Wilpert and David Mayo-Muñoz — Bac­te­ria, just like hu­mans, are fre­quently in­fected by viruses (bac­te­rio­phages or phages). Phages repli­cate at the ex­pense of their hosts, killing them and the process re­leases hun­dreds of prog­eny into the en­vi­ron­ment to com­plete their life cy­cle. The end of bac­te­ria? Not so fast!

  • Back from the ISS – A Snip­pet

    by Christoph — In this blog, we don't usu­ally con­sider big, mul­ti­cel­lu­lar an­i­mals, say, arthro­pods like Drosophila, ne­matodes like C. el­e­gans, or tardi­grades like Hypsi­bius du­jar­dini. We do make ex­cep­tions for the hosts of mi­cro­bial en­dosym­bionts. Why al­low the 'wa­ter bears' (tardi­grades)? With body lengths of 0.5‒1 mm, they are a bit too big to be hon­orary mi­crobes but...