High   ROS   levels   in   tumor   cells   as   a   basis  for  cancer  therapy

High   ROS   levels   in   tumor   cells   as   a   basis  for  cancer  therapy  

George  Liao   MPS  –  2060442    

Master’s  Thesis     Under  supervision  of   Dr.  Inge  A.M.  de  Graaf   Department  of  Pharmacokinetics,    

Toxicology  and  Targeting     Second  assessment  by   Dr.  Angela  Casini   Department  of  Pharmacokinetics,   Toxicology  and  Targeting     Oct  17  2014  

Abstract  

Cancer  is  the  result  of  aberrant  regulation  of  growth  and  death  in  cells.  Due  to  its   persistent   nature,   current   therapies   are   usually   unselective   and   highly   toxic.   Thus,   therapies  may  frequently  result  in  severe  adverse  effects.    

Reactive  oxygen  species  are  reactive  molecules  that  play  a  role  in  cellular  signaling   and   are   formed   as   side   products   in   physiological   processes.   Oxidative   stress   may   result  in  carcinogenesis  and  further  cancer  physiology  maintenance.  Cancer  cells  are   known  to  exhibit  significantly  elevated  ROS  levels.  Many  factors  may  contribute  to   this   increase;   regardless,   cancer   cells   regulate   their   ROS   levels   rigorously   as   high   levels  may  also  be  toxic  to  them.  

Recent   developments   in   cancer   therapy   have   thus   focused   on   finding   ways   to   use   this  high  ROS  level  in  cancer  to  therapeutic  advantages.  In  this  essay,  three  distinct   approaches   to   use   ROS   have   been   discussed   for   treatment   options:   namely,   by   generating   additional   ROS   molecules;   by   impairing   the   antioxidant   defenses   in   the   cancer  cells;  or  by  developing  prodrugs  that  are  activated  by  ROS  molecules.  

The  current  ROS-­‐generating  compounds  and  antioxidant  modulating  compounds  are   cancer   selective   because   it   only   pushes   the   ROS   concentration   over   the   toxic   threshold   in   cancer   cells;   the   prodrug   approach   may   be   truly   selective   because   it   requires   the   aberrant   high   ROS   levels   in   cancer   cells   to   activate   it.   All   three   approaches   are   still   mostly   in   their   infancy,   yet   many   of   these   compound   classes   have  shown  promising  results  in  preclinical  studies.  However,  many  factors  such  as   clinical  efficacy  have  yet  to  be  examined  for  the  majority  of  the  compounds.  A  viable   strategy  could  be  to  pair  these  ROS-­‐modulating  compounds  with  each  other  or  with   existing  chemotherapeutics.  

INTRODUCTION   4  

REACTIVE  OXYGEN  SPECIES  IN  CANCER   9  

Elevated  ROS  levels   9  

ROS  levels  beyond  the  threshold   11  

REACTIVE  OXYGEN  SPECIES  AS  A  POTENTIAL  THERAPEUTIC   13  

ROS  Generation   13  

Motexafin  gadolinium   13  

Hispidin   14  

Hirsutanol  A   15  

Methyl  3-­‐(4-­‐nitrophenyl)  propiolate   16  

ROS  generation  by  secondary  mechanisms   17  

Antioxidant  modulation   18  

Arsenic  trioxide   18  

Buthionine  sulfoximine   19  

Mangafodipir   21  

β-­‐Phenethyl  isothiocyanate   22  

Antioxidant  modulation  conclusion   23  

Prodrug  Approach   24  

Boronic  acids  and  esters   24  

Multiple  effectors   27  

CONCLUSION  AND  PROSPECTS   28  

Introduction  

Cancer  is  the  collective  name  for  the  group  of  diseases  that  results  from  aberrant  cell   growth   and   proliferation.   Specifically,   cancerous   cells   are   characterized   by   an   enhanced  rate  of  growth,  loss  of  function  and  the  potential  to  invade  healthy  tissue.¹   Neoplasms,  or  more  commonly  tumors,  are  the  result  of  aberrant  growth  of  tissue.  

However,  not  all  neoplasms  are  cancerous;  in  order  to  be  considered  cancerous,  the   neoplasm  has  to  be  invasive  and  deleterious  to  its  surrounding  tissue.  

As  cancer  is  a  whole  spectrum  of  diseases,  there  is  not  one  single  causative  agent,   nor   one   single   underlying   mechanism.   Even   with   the   same   cancer,   diverse   mechanisms  can  be  the  cause.  However  in  all  cases  for  cancer  to  develop,  the  cells   have   to   acquire   mutations   that   allow   growth-­‐regulation   bypass   and   apoptosis   evasion,  both  of  which  are  normally  under  strict  supervision.²    

Key  mutations  can  be  inherited  through  genetics  or  acquired  from  external  factors.  

Many  external  factors  are  implicated  with  carcinogenesis,  one  of  which  are  reactive   oxygen   species.   These   molecules   are   able   to   oxidize   cellular   macromolecules   and   DNA,   potentially   resulting   in   damage   and   mutations   to   proteins   and   genes   that   usually   manage   cell   growth,   proliferation   or   apoptosis.³  Cancer   arises   when   repair   mechanisms   fail   for   these   so-­‐called   oncogenes   and   tumor   suppressor   genes;   for   instance,   resulting   in   constitutive   activity   of   growth   factor   receptors   or   loss   of   function   of   the   DNA-­‐repairing   and   pro-­‐apoptotic   p53   protein.²   Research   has   determined  that  many  cancer  types  show  aberrant  activity  for  the  same  proteins.²    

Reactive  oxygen  species  (ROS)  are  inherently  formed  during  physiological  processes,   such  as  for  the  catalysis  of  nutrients  in  peroxisomes⁴  or  extermination  of  pathogens   by   immune   cells,³   but   also   as   byproducts   during   mitochondrial   oxidative   phosphorylation,⁵  enzymatic   protein   folding⁶  or   metabolism   of   xenobiotics.⁷   Additionally,   ROS   may   also   play   redox   signaling   roles.⁸  These   amounts   of   ROS   are   unlikely  to  cause  cancer  as,  in  addition  to  DNA  repair  mechanisms,  healthy  cells  can   tolerate   ROS   because   of   antioxidant   systems   to   ‘disarm’   them.³   Still,   a   delicate   balance  exists  between  the  formation  and  elimination  of  ROS  in  all  cells.    

Disturbance  in  this  balance  will  result  in  accumulation  of  ROS.  The  resulting  state  is   then   called   oxidative   stress.   It   is   this   oxidative   stress   that   could   initiate   cancer   by   causing  mutations;  the  high  amounts  of  ROS  could  prove  to  be  too  overwhelming  for   the   cellular   repair   mechanisms.   Following   cancer   initiation,   ROS   levels   have   been   determined   to   be   constitutively   elevated   in   cancer   cells.⁹  This   enables   ROS   to   also   play  an  important  role  in  cancer  promotion  and  further  exacerbation;¹⁰  this  topic  has   already  been  covered  by  many  reviews,  thus  it  lies  beyond  the  scope  of  the  current   essay.   Interestingly   though,   persistent   high   levels   of   these   molecules   can   also   be   deleterious  to  cancer  cells.^{11,  12}  In  order  to  survive  and  thrive  under  this  heightened   oxidative  stress,  various  antioxidant  systems  were  also  discovered  to  be  upregulated   in   cancer;   however   some   physiological   antioxidants   were   observed   to   be   diminished.11,  13,  14    

It   can   thus   be   concluded   that   the   regulation   of   ROS   in   cancer   is   a   very   complex   matter;  this  intricacy  is  why  ROS  implications  in  cancer  are  not  fully  clear.  However,   the   given   fact   that   a   threshold   for   ROS   level   exists   in   cancer   could   prove   to   be   a   valuable   anti-­‐cancer   strategy.   By   using   the   high   levels   to   our   advantage   or   further   manipulating  the  balance,  drugs  could  potentially  be  developed  that  are  activated  by   high   ROS   levels   (prodrugs)¹⁵  or   induce   cytotoxicity   by   generating   additional   ROS   molecules.7,  16    

At  present,  only  few  anti-­‐cancer  drugs  are  truly  selectively  targeting  the  neoplasms.  

Most   drugs,   however,   are   non-­‐selective   and   attempt   to   target   cancer   cells   by   exploiting   their   characteristic   rapid   growth   by   inhibiting   cell   division.¹⁷  As   a   result,   healthy   cells   frequently   sustain   collateral   damage,   which   is   a   significant   problem   associated  with  chemotherapy.  It  is  thus  of  the  utmost  importance  to  develop  novel,   selective   drugs   to   challenge   the   unmet   medical   need.   As   ROS   levels   differ   significantly   in   healthy-­‐   and   cancer   cells,   drugs   that   make   use   of   this   altered   physiology  may  potentially  be  used  as  targeted  therapy.  

An  illustration  summarizing  the  different  effects  of  varying  ROS  levels  is  presented  in   Figure  1,  as  obtained  from  the  review  on  ROS  in  cancer  by  Liou  and  Storz.¹⁸  

The  aim  of  this  essay  is  to  explore  how  the  high  ROS  levels  in  cancer  cells  may  be   used  to  develop  targeted  cancer  therapies.  As  ROS  are  inherently  toxic  molecules,   cancer  cells  are  thus  at  a  heightened  risk  for  ROS  damage.  In  order  to  understand  the   possibilities   of   a   treatment   with   ROS,   a   general   overview   of   physiological   ROS   maintenance   and   functioning   will   first   be   given,   followed   by   the   discussion   of   the   possible  underlying  mechanisms  of  the  constitutive  oxidative  stress  in  cancer  cells.  

Finally,  the  compounds  that  utilize  the  high  ROS  levels  as  part  of  their  mechanism  of   action  will  be  reviewed.  

Figure   1   The   concentration   of   reactive   oxygen   species   in   a   cell   determines   its   effect.   Low   ROS   levels   are   physiological   as   they   are   formed   during   biochemical   processes  and  may  take  part  in  cell  signaling.  Elevated  ROS  concentrations  result   in   oxidative   stress,   which   could   cause   carcinogenesis,   cancer   promotion   and   exacerbation.   Beyond   a   threshold,   ROS   becomes   toxic,   even   to   the   tumor,   and   could  be  used  in  a  new  anti-­‐cancer  approach.  

Reactive  oxygen  species  

Reactive   oxygen   species   is   the   collective   name   for   oxygen-­‐containing   reactive   molecules.   Despite   their   toxicity,   these   molecules   are   ubiquitous   in   all   living   organisms.  Figure  2  shows  three  of  the  most  common  ROS  species,  as  well  as  how   they  are  related  to  the  reduction  of  oxygen  to  water.  

Of   the   three   ROS   molecules,   the   hydroxyl   radical   is   the   most   reactive.   Due   to   its   extremely  short  half-­‐life  (nanoseconds),  HO^•  will  react  with  any  structure  close  to  its   site  of  formation.³  Hydrogen  peroxide,  on  the  other  hand,  is  the  least  reactive  of  the   three;  however,  in  no  way  does  that  mean  it’s  the  least  harmful:  due  to  its  neutral   charge  and  longer  half-­‐life  (milliseconds),  H2O2  has  the  ability  to  diffuse  away  from   its   source   and   even   cross   cellular   membranes.³   It   may   then   react   directly   with   cellular  targets,  but  it  can  also  be  reduced  to  the  highly  reactive  HO^•  by  transition   metals  in  the  Fenton  reaction,  as  presented  in  Figure  3.³  

Many  antioxidant  systems  exist  to  attenuate   ROS   molecules;   the   three   most   common   antioxidant   enzymes   and   their   reaction   are   shown  Figure  4.  Superoxide  dismutase  (SOD)   converts   2   moles   of   O2•-­‐   to   the   less   reactive   H2O2.  To  prevent  conversion  of  H2O2  to  HO^•,   the   enzymes   catalase   and   glutathione   peroxidase   (GS-­‐Px)   are   responsible   for   its   conversion   to   water   and   oxygen.³   Oxidative   stress  arises,  with  its  associated  toxicity,  when  a   disturbance  in  the  balance  takes  place  by  either   an  overproduction  of  ROS,  or  a  depletion  of  the  antioxidant  defenses.  

O₂ e O₂ e H₂O₂ e HO e H₂O Oxygen Superoxide anion

radical

Hydrogen peroxide

Hydroxyl

radical Water

O₂

2 + 2H⁺ Superoxide Dismutase H₂O₂ + O₂ H₂O₂ + RH₂ Catalase R + 2 H₂O

2 H₂O₂ 2 H₂O + O₂ 2 GSH + H₂O₂ Glutathione Peroxidase GS–SG + 2 H₂O

Catalase

Figure  2  The  stepwise  reduction  of  oxygen  to  water  yields  3  reactive  oxygen  species.³  

Figure   3   The   Fenton   reaction   involves   the   reduction  of   H2O2  to   the  highly   reactive   HO^•   (encircled)  and  a  hydroxyl  ion  by  a  transition   metal.   O₂^–•   may   drive   the   reaction   by   reducing  the  oxidized  metals.  

H₂O₂ HO + HO

Fe²⁺ Fe³⁺

O₂ O₂

Figure   4   The   schemes   of   the   reactions   catalyzed   by   the   three   major   antioxidant   enzymes.  

Superoxide  dismutase  catalyzes  the  conversion  of  2  moles  of  O2–•  to  H2O2  and  oxygen.  Catalase  is   able   to   further   reduce   H₂O₂   to   water;   in   peroxisomes,   the   enzyme   oxidizes   other   molecules   simultaneously   with   the   reduction   of   peroxide.   When   H2O2   accumulates,   catalase   can   switch   to   reducing  2  moles  of  peroxide  at  the  same  time.  Glutathione  peroxidase  is  also  able  to  reduce  H2O2   to  water  by  oxidizing  the  scavenger  glutathione  (GSH).  

Traditionally,  ROS  have  always  been  associated  with  oxidative  stress  and  its  potential   to  cause  damage  to  cellular  structures.  However  recently,  ROS  molecules  have  been   implicated  with  physiological  signaling  in  the  cell.⁸  Several  downstream  transduction   pathways   of   growth   factors   and   cytokines   seem   to   involve   ROS   production.   The   specific  mechanisms  are  still  unclear,  but  one  important  hypothesis  is  that  ROS  may   play   a   role   in   the   oxidative   modification   of   downstream   proteins,   for   instance   kinases  or  transcription  factors.⁸  

Under   physiological   conditions,   the   main   source   of   ROS   production   is   the   peroxisome  as  H2O2  are  generated  to  oxidize  and  catabolize  nutrients  by  catalases.⁴   However,  when  large  amounts  of  H2O2  accumulate  in  peroxisomes,  these  molecules   may  escape  by  diffusion  and  interact  with  other  cellular  structures.⁴  

The  ER  is  the  major  site  for  mRNA  translation  and  post-­‐translational  modifications   such  as  disulfide  bond  formation.  The  protein  that  is  responsible  for  bond  formation   incidentally  forms  H2O2  in  the  process  as  a  side  product.^{3,  6}    

NADPH  oxidase,  which  is  primarily  located  on  the  cell  membranes  of  immune  cells,   produces  O2–•  to  exterminate  invading  pathogens.¹⁹  The  O2–•  molecules  are  released   at  once  during  respiratory  burst,  but  as  ROS  molecules  are  non  selective,  this  release   has   the   potential   to   damage   other   cells   as   well;   hence   the   association   of   inflammation   with   cancer.   Other   isoforms   of   NADPH   oxidase   are   present   on   non-­‐

immune   cells   and   may   be   involved   with   the   production   of   ROS   for   signaling   purposes.⁸  

Redox  active  enzymes,  such  as  the  CYP  enzymes,  are  important  for  the  metabolism   of   xenobiotics.   However,   electrons   may   leak   out   in   the   process   and   create   ROS   molecules;   this   process   is   called   uncoupling.³   Additionally,   depending   on   the   compound,   CYP   enzymes   may   induce   ROS   promotion   through   redox   cycling:   the   compound  is  first  reduced  by  CYP,  after  which  the  electron  is  passed  on  to  oxygen,   thus  making  it  a  substrate  for  CYP  again  and  so  on.³  

The  electron  transport  chain  (ETC),  located  on  the  mitochondrial  inner  membrane,  is   a   very   important   site   of   ROS   generation;⁵   in   addition   to   physiological   ROS   production,  it  is  also  the  main  site  for  exogenous  induction.³  The  ETC  is  an  intricate   network   of   proteins   that   is   responsible   for   the   ATP-­‐generating   oxidative   phosphorylation.  As  the  name  suggests,  electrons  are  transported  over  the  network   from  donating  cofactors  to  oxygen  molecules  at  the  end  to  form  water.  This  process   of  electron  transport  is  prone  to  leakage  however,  which  means  that  oxygen  may  be   prematurely   reduced,   resulting   in   O2–•.⁵   Escaped   ROS   molecules   can   then   interact   with   structures   of   the   ETC,   mitochondrial   macromolecules   (such   as   mitochondrial   DNA  (mtDNA))  or  even  other  structures  within  the  cell.  Due  to  the  potential  for  large   amounts   of   O2–•  to   form,   mitochondria   have   their   own   SOD   isoform   to   further   reduce   the   molecules;   the   mitochondrial   SOD   binds   manganese   and   is   thus   called   MnSOD.⁵  

ROS   can   thus   be   generated   at   many   different   sites;   luckily,   under   physiological   conditions,   the   antioxidant   systems   are   amply   capable   to   neutralize   the   ROS   molecules  before  irreparable  damage  can  be  inflicted.  However,  if  the  homeostasis   is   affected,   ROS   may   accumulate   and   potentially   initiate   cancer   by   either   directly   damaging  the  genome  on  oncogene  or  tumor  suppressor  gene  positions  or  indirectly   by  interacting  with  the  redox-­‐susceptible  signaling  pathways.  

Reactive  oxygen  species  in  cancer  

Following   cancer   initiation,   research   has   determined   that   basal   ROS   levels   are   constitutively   elevated   in   cancer   cells.⁹   The   exact   mechanism   for   this   shift   in   ROS   balance   is   still   unclear,   as   many   factors   could   potentially   contribute.   Thus,   the   factors  that  can  induce  ROS  molecules  in  general  will  be  presented  in  this  chapter,   followed  by  a  discussion  on  how  cancer  cells  may  adapt  to  survive  under  oxidative   stress.  

Elevated  ROS  levels  

As   cancer   is   a   diverse   disease   with   many   forms   of   expression,   many   mechanisms   could  therefore  contribute  to  the  elevated  basal  ROS  levels.  The  originating  location   of  the  neoplasm  is  also  important  as  differentiated  cells  have  different  expressions   of  cellular  constituents  and  cells  may  thus  function  differently  even  as  cancer  cells.  

As   previously   described,   the   mitochondria   are   an   important   source   of   ROS.   Under   normal   circumstances,   the   electron-­‐leakage   from   the   ETC   is   easily   manageable   by   antioxidants.   However,   this   balance   can   be   seriously   affected   following   mitochondrial  damage;  ROS  formation  may  namely  be  significantly  increased  when   the   ETC   or   mitochondrial   DNA   become   damaged.²⁰  Direct   damage   to   structures   of   the  ETC  could  result  in  decreased  efficiency  of  electron  transport,  which  may  cause   an   increased   leakage   of   electrons   and   thus   an   increased   premature   oxygen   reduction.  Damage  to  the  mitochondrial  DNA  results  in  an  impaired  ETC  as  well,  as   13   of   the   100   ETC-­‐proteins   are   encoded   by   mtDNA.²⁰   Due   to   the   convenient   proximity  of  the  mitochondrial  structures  to  the  site  of  ROS  production,  ROS  is  thus   an  important  factor  for  its  own  amplification.    

In  parallel  with  the  altered  ROS  levels,  research  has  determined  that  cancer  cells  may   also   have   aberrant   levels   and   regulation   of   antioxidant   systems,   which   could   potentially  contribute  to  the  shift  in  ROS  regulation  towards  higher  levels.²¹  Although   one  must  keep  in  mind  that  this  is  highly  cancer  specific  as  cellular  constituents  in   cancer,   including   antioxidants,   depend   on   the   healthy   cell’s   physiological   function.  

However,  in  their  review,  Oberley  and  Oberley  have  noticed  that  cancers  involving   the  lungs,  kidney  and  prostate  often  show  diminished  levels  of  catalase  and  GS-­‐Px,   which  will  result  in  the  accumulation  of  H2O2.²¹  On  the  other  hand,  the  mitochondrial   MnSOD   is   diminished   in   many   cancer   forms,   causing   accumulation   of   O2–•.²¹   Aberrant  regulation  of  antioxidants  can  thus  potentially  contribute  to  accumulation   of  ROS,  however  this  aspect  is  highly  cancer  specific  and  the  size  of  its  role  differs.    

One   of   the   main   characteristics   of   cancer   is   the   uninhibited   proliferative   power   of   the  cells.  Oncogenes,  such  as  Ras,  Bcr-­‐Abl  and  c-­‐Myc,  are  fundamental  for  this  aspect   as   many   cancer   forms   express   overactivity   of   their   encoded   proteins,   which   have   functions   associated   with   inducing   cell   growth   and   division.²   Oncogene   activity   of   the  aforementioned  Ras,²²  Bcr-­‐Abl²³  and  c-­‐Myc²⁴  have  all  been  associated  with  ROS   production;⁸   thus   in   the   case   of   cancer   with   hyperactive   oncogenes,   significant   amounts   of   ROS   may   be   produced   during   normal   cancer   functioning.   As   with  

physiological  signaling  involving  ROS,  the  exact  reason  for  the  pair-­‐wise  production   of  ROS  with  oncogene  activation  is  also  unclear.  The  ROS  molecules  could  potentially   promote  cancer  growth  by  causing  additional  genomic  instability;  or,  it  could  simply   perform  its  physiological  function  of  oxidative  modification  of  downstream  proteins,   including  transcription  factors.  

As   a   cancer   cell   grows   rapidly   and   excessively,   eventually   even   beyond   its   normal   tissue   structures,   it   may   end   up   outgrowing   its   physiological   blood   supply.   As   a   result,  the  blood  vessels  may  not  be  able  to  supply  all  areas  of  the  neoplasm  with   sufficient   oxygen;   hypoxia   could   thus   become   a   problem   in   these   areas.   H2O2   has   been   implicated   with   the   signaling   for   angiogenesis   by   stabilizing   the   transcription   factor  hypoxia-­‐inducible  factor-­‐1  (HIF-­‐1).²⁵  These  ROS  molecules  originate  from  the   mitochondria   and   are   induced   under   hypoxic   conditions:   under   low   oxygen   conditions,  the  ETC  is  impaired  which  may  cause  more  electrons  to  leak  prematurely   than   normal.²⁶  The   ‘leaked’   electrons   will   reduce   the   scarce   amounts   of   oxygen   available   to   form   O2–•,   which   is   subsequently   reduced   by   MnSOD   to   H2O2.²⁵   As   hypoxia   is   a   common   condition   in   tumor   cells,   it   may   very   well   be   one   of   the   contributors  to  the  elevated  ROS  levels.    

Recently,   a   group   of   researchers,   led   by   El   Sayed,   have   hypothesized   that   the   Warburg   effect   may   be   a   contributing   factor   to   the   high   basal   ROS   levels.²⁷  The   Warburg  effect  is  the  phenomenon  that  cancer  cells  show  a  high  level  of  pyruvate   fermentation  rather  than  oxidative  metabolism  through  the  Krebs  cycle,  regardless   of  the  level  of  oxygen  available.²⁸  Krebs  cycle  intermediates²⁹  have  been  implicated   with  antioxidant  activities.  El  Sayed  suggests  that  due  to  the  tendency  of  cancer  cells   to  generate  energy  primarily  by  fermentation,  less  Krebs  cycle  intermediates  will  be   available   to   aid   in   modulating   ROS   levels.   Additionally,   the   Krebs   cycle   produces   NADH,  which  is  indirectly  used  to  reduce  oxidized  glutathione.  While  El  Sayed  and  his   coworkers   have   a   point   that   fermentation   may   result   in   diminished   antioxidant   capacity  of  Krebs  intermediates,  this  can  only  be  a  small  contributor  in  the  elevated   ROS  levels.  The  Krebs  cycle  intermediates  are  namely  weak  antioxidants  and  require   concentration  levels  exceeding  the  physiological  values  to  have  the  same  potency  as,   for  instance,  glutathione;²⁹  this  is  due  to  their  postulated  mechanism  as  they  likely   exert   their   antioxidant   properties   by   chelating   reduced   transition   metals.²⁹   Furthermore,   lactic   acid,   the   resulting   compound   from   pyruvate   fermentation,   is   known   to   exhibit   antioxidant   properties   as   well.³⁰  The   only   strong   argument   in   El   Sayed  et  al’s  favor  is  that  the  intermediate  fumarate  has  recently  been  associated   with   activating   the   important   antioxidant   transcription   factor   Nrf2.³¹  On   the   other   hand,  El  Sayed  has  not  taken  into  account  that,  in  addition  to  fermentation,  cancer   cells   still   undergo   oxidative   metabolism   via   the   Krebs   cycle.³²  Thus,   it   can   be   concluded   that   the   Warburg   effect   is   unlikely   the   main   cause   for   –or   even   a   big   contributor  to–  the  elevated  ROS  levels.    

Telomeres  are  the  non-­‐coding  nucleotides  at  the  ends  of  chromosomes  that  protect   the  genome  from  losing  ‘information’  during  replication  of  the  lagging  DNA  template   strand.³³  Telomere   dysfunction,   which   is   the   shortening   of   telomeres,   may   cause   genomic   instability   and   thus   the   initiation   of   cancer.   Interestingly,   telomere  

dysfunction   has   recently   also   been   linked   with   mitochondrial   fitness.³⁴  The   shortening   of   telomeres   may   namely   induce   the   tumor   suppressor   protein   p53,   which  is  involved  in  numerous  regulatory  pathways.  The  transcriptional  coactivators   PGC-­‐1α   and   PGC-­‐1β,   which   are   important   regulators   for   mitochondrial   biogenesis   and  functioning  –including  antioxidant  generation,  are  associated  with  upstream  p53   signaling;  telomere  dysfunctioning  will  result  in  the  inhibition  of  these  proteins.  As  a   result,   the   inhibition   of   PGC   coactivators   by   telomere   dysfunctioning   will   result   in   mitochondrial  impairment,  which  will  result  in  increased  leakage  of  electrons  from   the  ETC  and  thus  ROS  generation.³⁴  It  is  however  important  to  know  that  p53  has  to   function  for  this  to  contribute  to  extra  ROS  generation,  which  is  not  always  the  case.  

The   tumor   suppressor   protein   p53   is   a   very   important   factor   in   physiological   cell   functioning;  it  is  namely  responsible  for  preventing  the  formation  of  cancer,  hence   the  name  ‘tumor  suppressor’.  Furthermore,  p53  may  also  regulate  the  transcription   of  antioxidants.⁷  Its  main  functions  include  the  activation  of  DNA  repair  mechanisms,   the   suspension   of   mitosis   and,   in   case   of   irreparable   damage,   the   induction   of   apoptosis.³⁵  Many  cancers  are  associated  with  acquired  mutations  that  result  in  the   loss  of  functional  p53  protein.³⁵  This  loss  of  function  can  lead  to  a  vicious  cycle  of   genomic   instability;   namely,   less   antioxidants   will   be   synthesized   and   ROS-­‐induced   DNA  damage  will  not  be  repaired,  which  will  lead  to  accumulation  of  damage  and   further   dedifferentiated   cells   following   cell   division.   In   their   turn,   these   cells   are   prone  to  produce  even  more  ROS  and  so  on.³⁵    

ROS  levels  beyond  the  threshold  

Despite   the   importance   of   ROS   for   cancer   initiation,   promotion   and   progression,   they   are   still   inherently   noxious   molecules   with   the   potential   to   kill   cancer   cells.  

While   the   basal   ROS   levels   are   significantly   elevated   in   cancer   physiology,   the   concentration   is   still   contained   below   the   toxic   threshold   (Figure   1).   Cancer   cells   manage   to   survive   under   these   conditions   by   altering   the   expression   of   their   antioxidant  defenses  or  the  proteins  involved  with  survival  and  death.    

As   described   above,   the   antioxidant   expression   of   cancer   cells   has   already   been   altered.   Many   transcription   factors   involved   in   the   expression   of   antioxidant   enzymes   are   known   to   be   redox   sensitive.¹⁰   High   ROS   levels   could   therefore   be   sensed   by   these   transcription   factors   and   the   expression   of   antioxidants   will   subsequently  be  induced.  Additionally,  oncogenes  may  also  induce  the  synthesis  of   various   antioxidants,   which   could   also   be   linked   to   their   production   of   ROS   molecules.   Cells   with   high   Ras   activity   have   been   determined   to   have   high   peroxiredoxin  concentration,³⁶  which  is  also  an  enzyme  that  is  responsible  for  H2O2   reduction   to   water,   while   c-­‐Myc   activity   is   associated   with   induction   of   GSH   synthesis.³⁷  

In  addition  to  upregulating  the  synthesis  of  antioxidants,  cancer  cells  may  also  cope   with   the   high   ROS   levels   by   interfering   with   apoptosis.   The   redox-­‐sensitive   transcription   factors   can   namely   also   enhance   the   synthesis   of   anti-­‐apoptotic   factors¹⁰  or  diminish  the  activity  of  pro-­‐apoptotic  factors.³⁸  However,  when  the  ROS   level  somehow  manages  to  surpass  the  closely  guarded  toxic  threshold,  cancer  cell  

cycle   arrest   and   cell   death   can   take   place,   mediated   by   ROS.   ROS   may   attenuate   cancer   by   either   inducing   apoptosis   and   senescence   via   pro-­‐apoptotic   proteins,   or   damaging   the   cancer   cells   to   the   point   of   irreparable   damage.^12,  18,  39  The   various   mechanisms  for  ROS-­‐mediated  cancer  attenuation  will  be  elucidated  in  the  following   section  for  each  drug,  when  available.  

The  dual  action  of  ROS  may  seem  paradoxical,  however  recently,  research  has  found   evidence  that  O2–•  and  H2O2  may  have  differential  activity  in  cancer  cells.⁴⁰  In  their   review  article  Pervaiz  and  Clement  present  evidence  that  high  O2–•  levels  are  linked   with  apoptosis  resistance  in  cancer  cells,  as  SOD  isoforms  seem  to  be  downregulated   whereas  the  anti-­‐apoptotic  Bcl-­‐2  is  up-­‐regulated.  On  the  other  hand,  high  H2O2  levels   are  associated  with  apoptosis.  This  may  potentially  be  due  to  hydrogen  peroxide’s   ability  to  diffuse  away  from  its  source  and  interact  with  cellular  structures,  such  as     the   mitochondrial   membrane   or   caspases.   Interaction   with   the   membrane   could   result  in  an  alteration  of  its  permeability,  resulting  in  the  release  of  cytochrome  C;  

interaction   with   caspase   proteases   could   result   in   its   activation.   Apoptosis   will   be   induced  in  both  these  scenarios.  In  any  case,  the  exact  mechanisms  are  still  unclear,   however  evidence  suggests  that  high  H2O2  levels  may  act  as  pro-­‐apoptotic  factors.  

Reactive  oxygen  species  as  a  potential  therapeutic  

Equipped  with  the  knowledge  that  cancer  cells  have  significantly  higher  ROS  levels   than   healthy   cells,   viable   targeted   treatments   could   thus   be   aimed   at   turning   an   important  carcinogenic  to  our  own  advantage.  

Indeed,  researchers  have  found  three  distinct  ways  to  fight  cancer  cells  through  ROS-­‐

related  mechanisms.  Namely,  cancer  cells  can  be  killed  by  generation  of  additional   ROS;  or  ROS  levels  can  be  further  raised  by  diminishing  their  antioxidant  capacity;  

and  finally,  the  high  ROS  levels  could  be  employed  to  activate  prodrugs  specifically  in   cancer  cells.  

This  section  of  the  essay  will  elaborate  on  the  mechanisms  of  action  by  which  the   three   ‘classes’   of   ROS-­‐related   anti-­‐cancer   drugs   work.   The   three   classes   will   be   divided  in  three  separate  paragraphs  and  drugs  in  each  class  will  include  its  proof  of   concept  in  the  form  of  experimental  in  vitro  or  in  vivo  data,  if  available.  

ROS  Generation  

Cancer  cells  rigorously  maintain  their  ROS  levels  below  the  toxic  threshold.  As  ROS   levels  are  constitutively  elevated  significantly  in  cancer  cells  and  antioxidant  systems   have  already  been  enhanced  in  order  to  survive,  additional  generation  of  ROS  could   thus  be  able  to  tip  the  balance  towards  toxicity  and  cancer-­‐killing.  Cancer  selectivity   in  this  case  is  based  on  the  ability  of  normal  cells  to  easily  deal  with  the  additional   ROS.  

Motexafin  gadolinium  

Motexafin   gadolinium,   commercially   known   as   Xcyclin,  is  a  synthetic  porphyrin,  which  is  heterocyclic   macrocycle. ⁴¹  After   showing   high   potential   in   preclinical   studies   for   the   treatment   of   lung   cancer   brain   metastases,   the   drug   was   ultimately   still   disapproved   by   the   FDA   (reason   not   specified);  

researchers   are   now   investigating   Xcyclin   for   non-­‐

small  cell  lung  cancers.⁴²      

Xcyclin  can  be  listed  under  both  the  ROS  Generation   or   Antioxidant   modulation   paragraphs   as   the   compound  was  found  to  deplete  the  reducing  agent   NADPH   and   the   antioxidants   ascorbate   (vitamin   C)   and   GSH,   as   well   as   induce   ROS   generation   through   redox   cycling.⁴¹   The   resulting   high   levels   of   ROS   are  

suspected   to   interact   with   the   mitochondrial   membranes,   causing   release   of   cytochrome  c  and  subsequent  induction  of  apoptosis.⁴¹  

In   both   in   vitro   and   in   vivo   studies,   researchers   found   that   the   compound   is   selectively   taken   up   by   cancer   cells.⁴³  This   selective   localization   was   expected,   as   natural  porphyrins  are  known  to  be  taken  up  by  cancer  cells,  although  the  reason  for  

N

N Gd HO

N N

N OH

O

O

O O

O

O

O O

O O O

O

Structure  1  Motexafin  gadolinium  

this   is   still   unclear.⁴⁴  The   subsequent   clinical   studies   have   determined   that   Xcyclin   could  synergistically  enhance  the  activity  of  other  cancer  treatments.  In  the  case  of   brain   metastases,   the   drug   improved   the   neurological   functioning   of   the   patients   over   exclusively   brain   radiation;   however,   no   improvement   in   anti-­‐cancer   activity   was  found.⁴⁵  

Hispidin  

Hispidin   is   a   natural   compound   obtained   from   a   medicinal  mushroom  used  for  centuries  in  traditional   Asian   medicine.⁴⁶  Included   in   its   indications   was   the   treatment   for   cancer.   Researchers   have   recently   discovered   that   Hispidin   induces   apoptosis   in   colon   cancer  cells  through  ROS  level  manipulation.⁴⁷    

As  the  discovery  of  Hispidin’s  activity  is  recent,  only  fundamental  proof  of  concept   experiments  were  thus  far  conducted  in  vitro.  Firstly,  colon  cancer  cells  were  dose-­‐

dependently  killed  by  Hispidin  as  determined  by  Annexin  V  staining.⁴⁷  Pre-­‐incubation   of   the   cells   with   a   ROS   scavenger   showed   that   the   activity   of   Hispidin   was   diminished,   thus   proving   ROS   the   likely   executioner   for   Hispidin.   Subsequently,   during   gene   expression   studies,   the   researchers   found   that   Hispidin,   through   ROS,   induced   the   expression   of   p53   protein   and   its   downstream   pro-­‐apoptotic   factors,   whereas   anti-­‐apoptotic   factors   were   downregulated.   Even   more   interestingly,   the   researchers   discovered   that   in   p53-­‐null   cells,   extrinsic   apoptotic   factors   were   induced,  thus  effectively  bypassing  the  necessity  of  p53  mediation.⁴⁷  Figure  6  shows   the  protein  expression  over  the  Hispidin  concentrations.  

Judging   from   the   in   vitro   data,   Hispidin   seems   to   be   a   compound   with   great   potential.   The   question   however   remains   whether   the   same   promising   results   are   expressed   in   vivo.   Furthermore,   the   source   of   the   ROS   molecules   is   still   unclear,   however   judging   by   the   structure,   a   quinone-­‐related   redox   cycling   might   be   its   mechanism.   On   the   other   hand,   cancer   selectivity   still   has   to   be   proven   as   the   researchers  have  not  reported  the  effect  of  hispidin  in  healthy  cells.  

HO

HO O

O OH

Structure  2  Hispidin  

Figure  5  Gene  expression  study  carried   out  on   rat   colon  cancer  cells  (CMT-­‐93).  Bax  is  responsible   for   mitochondrial   pore   opening   and   cytochrome   C   release.   Bcl-­‐2   is   an   inhibitor   of   apoptosis.   As   can   be   seen  in  (A),  increase  of  Hispidin  concentration  results  in  the  decrease  of  Bcl-­‐2  and  increase  of  Bax  and   p53.  In  (B),  Hispidin  induces  the  activation  of  both  caspases  1  and  8,  as  well  as  the  death  receptor  3.  

Furthermore,  the  cleavage  of  Parp  signals  the  start  of  apoptosis.  These  proteins  are  associated  with  the   extrinsic  apoptosis  pathway.  

Hirsutanol  A  

Hirsutanol   A   (HirA)   is   a   natural   product   obtained   from   fungi   that   has   recently   been   identified   to   show   promising  anti-­‐cancer  activity;  apoptosis  can  be  induced   in   breast   cancer   cells,   colon   cancer   cells   and   hepatocellular   carcinomas.^{48,  49}  By   probing   the   breast-­‐  

and   colon   cancer   cells   that   were   exposed   to   HirA   with  

Annexin   V,   researchers   were   able   to   determine   that   the   compound   induced   apoptosis  in  a  dose-­‐dependent  manner.  In  addition,  experiments  with  colon  cancer   xenografts  in  mice  show  that  HirA  is  also  able  to  attenuate  tumor  growth  by  half  in   vivo  (Figure  6).⁴⁹  The  researchers  did  not  report  on  any  observed  side  effects.  

HirA   induces   apoptosis   by   generating   ROS   that   interacts   with   the   mitochondrial   membrane  resulting  in  the  release  of  cytochrome  c  to  the  cytosol;  cytochrome  c  is   then   able   to   activate   the   caspases,   the   apoptotic   executioners.   Inhibition   of   Hirsutanol  A  activity  by  incubating  the  cell  lines  with  ROS  scavengers  supports  the   ROS-­‐based   apoptosis   idea.   The   source   for   the   extra   ROS   generated   by   HirA   is   still   unknown,  however  by  testing  HirA  on  ETC-­‐impaired  cell  lines,  researchers  discovered   that  apoptosis  is  still  induced  in  the  same  manner.  Thus,  the  conclusion  can  be  made   that  HirA  induced  ROS  stem  from  somewhere  else  than  the  mitochondria.    

Furthermore,   the   group   also   suggested,   by   using   fluorescent   oxidative   stress   indicators,  that  H2O2  was  the  main  molecule  to  be  generated,  as  the  CM-­‐H2DCF-­‐DA   probe   showed   significantly   more   fluorescence   than   the   DHE   probe   in   a   dose   dependent   fashion.⁴⁹   The   DHE   probe   is   specific   for   O2–•,   whereas   CM-­‐H2DCF-­‐DA   is   unspecific.  The  researchers  suggested  that  since  DHE  showed  minor  fluorescence,  it   is  most  likely  H2O2  that  is  causing  CM-­‐H2DCF-­‐DA  to  fluoresce,  as  O2–•  and  H2O2  are   the  most  frequently  occurring  ROS  species.    

As   result   of   the   HirA   induced   extra   ROS,   cells   were   found   to   induce   the   anti-­‐

apoptotic  JNK  signaling  pathway  to  diminish  the  effect  of  HirA;  thus,  treatment  with   HirA  paired  with  a  selective  JNK  signaling  inhibitor  should  further  improve  results.  As  

O

OH OH

Figure   6   The   effect   of   Hirsutanol   A   on   in   vivo   colon   cancer   xenografts   in   mice   plotted   with   normal   saline   (NS),   the   topoisomerase   1   inhibitor   HCPT   and   DMSO.   The   tumor   volume   is   approximately   cut   by   half   with   Hirsutanol   A.   The   experiment   was   terminated  following  measurements  on  day  21.  

Structure  3  Hirsutanol  A  

is  with  Hispidin,  the  researchers  did  not  report  on  the  cancer  selectivity  of  HirA  and   they  have  not  reported  the  effects  of  HirA  on  healthy  cells.  

Methyl  3-­‐(4-­‐nitrophenyl)  propiolate  

Methyl  3-­‐(4-­‐nitrophenyl)  propiolate  (NPP)  is  a  synthetic  ROS-­‐

inducing   compound   that   was   identified   in   high-­‐throughput   screening;  NPP  showed  potential  as  researchers  determined   that   it   could   induce   apoptosis   selectively   in   leukemia,   breast-­‐,   liver-­‐,   lung-­‐,   prostate-­‐,   colon-­‐,   skin-­‐   and   cervical   cancer  over  healthy  cells  in  vitro  (Figure  7).⁷    

Similar   to   HirA,   NPP   was   also   determined   to   induce  

apoptosis  in  a  dose-­‐dependent  fashion  (by  Annexin  V)  with  the  mode  of  action  being   the  release  of  cytochrome  C  to  the  cytosol.⁷  The  researchers  observed  that  ROS  was   generated  dose-­‐dependently  in  many  different  cell  lines,  including  non-­‐transformed   cells.⁷   Furthermore,   pre-­‐incubation   with   antioxidants   could   attenuate   cell   death;  

thus  proving  that  apoptosis  is  indeed  induced  by  the  generated  ROS.  Monitoring  the   ROS  levels  showed  that  the  molecules  were  induced  within  10  minutes,  peaked  at  30   minutes  and  started  diminishing  within  the  hour.⁷  

To  determine  the  source  of  the  ROS  molecules,  researchers  first  examined  NPP  on   ETC   impaired   cells   and   determined   that   no   significant   difference   in   effect   was   detectable.   Subsequently,   the   possibility   of   ROS   generation   by   CYP-­‐mediation   was   investigated  as  the  researchers  determined  that  the  reduced  propargyl  ester  of  NPP   might   behave   similarly   to   quinones   (Figure   9).   By   inhibiting   the   CYP   enzymes,   researchers   were   able   to   attenuate   the   ROS   formation   dose-­‐dependently;  

specifically   the   CYP3A4   inhibitors   showed   this   effect,   as   shown   in   Figure   8.   The   importance  of  the  propagyl  ester  was  found  when  carboxylesterases  in  hepatocyte   cell  lines  could  help  the  cell  withstand  NPP  exposure.⁷  

N⁺ O

-O

O O

Figure   8   ROS   levels   are   dose-­‐dependently   decreased   with   CYP3A4   inhibitors   TAO   (oleandomycin   triacetate)   and   CHL   (chloramphenicol).   The   CYP1A   inhibitors   ANF   (α-­‐

naphthoflavone)  lacked  this  effect.  

Structure   4   Methyl   3-­‐(4-­‐

nitrophenyl)   propiolate   (NPP).  The  propagyl  ester   is  colored  red.  

Figure   7   The   percentage   of   viable   cells   is   plotted   against  the  concentration  of    NPP  exposure.  As  can  be   seen,   the   non-­‐cancerous   Wi-­‐38   (lung)   and   LO2   (liver)   are   significantly   less   susceptible   to   NPP   exposure   compared  to  the  breast  carcinoma  (Hs578T)  and  breast   adenocarcinoma  (MDA-­‐MB-­‐468)  

In   terms   of   selectivity,   the   researchers   found   that   NPP   induces   apoptosis   preferentially  in  cells  with  high  ROS  levels  and  low  antioxidant  defenses;  leukemia   was   given   as   an   example.   In   line   with   this   discovery,   researchers   found   that   p53-­‐

dysfunctional  cells  are  more  susceptible  to  NPP.  This  is  a  profound  discovery  as  loss   of  functional  p53  is  found  in  the  majority  of  all  cancers  and  enables  tumors  to  form   and  proliferate  in  the  first  place,  as  well  as  evade  apoptosis.  In  this  case,  no  p53  will   be   able   to   upregulate   the   synthesis   of   antioxidants   following   NPP-­‐induced   ROS   generation.⁷    

Finally,  the  researchers  demonstrated  that  due  to  the  high  reactivity  of  ROS  species,   the  location  of  formation  might  be  more  important  than  the  quantity  of  ROS.  NPP  is   believed  to  induce  ROS  near  membranes,  as  CYP  enyzmes  are  membrane  associated.  

While  the  plasma  membrane  is  not  one  of  the  main  locations  for  CYP  enzymes,  it  is   known  to  have  CYP  enzymes.⁵⁰  The  researchers  postulated  that  the  enzymes  on  the   plasma   membrane   might   play   an   important   role   in   NPP   activity.   Near   the   plasma   membrane,   the   researchers   namely   found   that   ROS   molecules   could   inhibit   the   JAK/STAT   signaling   pathway,   possibly   by   oxidizing   the   cysteine   residues   of   the   JAK   tyrosine   kinase,   as   determined   by   Western   blotting;   as   a   result,   the   signaling   pathway  is  impaired  and  the  STAT-­‐promoted  expression  of  anti-­‐apoptotic  factors  is   attenuated.  A  summary  of  the  postulated  action  of  NPP  can  be  found  in  Figure  9.  

While   the   researchers   were   able   to   demonstrate   the   significant   effects   of   NPP   in   vitro,   the   question   remains   whether   its   potential   stays   intact   in   vivo.   The   main   concern  is  if  sufficient  CYP  enzymes  are  functional  in  cancer  cells,  especially  in  highly   dedifferentiated  forms,  and  if  they  are  at  the  correct  location  to  generate  the  ROS  in   order  for  NPP  to  exert  its  effect.  

ROS  generation  by  secondary  mechanisms  

Many  current  chemotherapeutics  are  known  to  generate  ROS,  either  as  a  secondary   mode   of   action   or   without   making   use   of   the   intrinsically   elevated   ROS   levels   in  

N⁺ O -O

O O

N⁺ O

-O C

O O CYP450

(reduced)

CYP450

(oxidized) O₂

O₂ H₂O₂

OH

Figure  9  The  proposed  mechanism  of  action  of  NPP.  NPP  can  undergo  redox  cycling  by  first  being  reduced   to   an   allenic   structure.   This  allenic   structure   then   passes   its  electron  to  oxygen   and  thus   reducing   it   to   superoxide  anion.  O₂^–•  can  further  be  reduced  to  other  ROS  molecules  and  these  may  cause  apoptosis  by   interacting  with  mitochondria  and  inhibiting  the  JAK/STAT  anti-­‐apoptotic  pathway.