The experiments presented indicate that members of the SSU processome sub-modules UTP-A and UTP-B continue to associate with early pre-ribosomes in strains disrupted in all tested r-proteins. Significantly, lack of assembly of r-proteins whose prokaryotic homologous proteins act according to in vitro
reconstitution experiments as primary rRNA binders in five of six prokaryotic SSU assembly trees, did not detectably reduce the association of these SSU processome components with early pre-ribosomes. Accordingly, robust incorporation of the SSU processome sub-module UTP-A into pre-ribosomes does neither depend on the presence of other tested SSU processome components 
nor on the presence of any of the tested r-proteins (see ). Altogether, this suggests that the UTP-A complex functions as transient primary binder in the hierarchy of eukaryotic SSU assembly.
Previous work indicated that the UTP-A sub-module in turn acts upstream of other SSU processome components including Noc4p 
. The data presented here indicate that rpS5 and other r-proteins of the head domain are still able to interact in vivo
to a certain extent with pre-ribosomes after inactivation of Noc4p. Nevertheless, establishment of more robust interactions of these r-proteins with rRNA required the presence of functional Noc4p. These observations reinforce the previous assumptions 
that the combined action of SSU processome components plays a crucial role in facilitating such specific assembly events, as the conversion of initial, weak r-protein - pre-18S rRNA interactions into a stable complex.
Several hypotheses can be taken into consideration on how SSU processome components might drive specific assembly events.
Establishment of robust interactions of most SSU r-proteins with 18S rRNA precursors correlates in normal conditions with SSU processome dependent cleavage events in the 5′-ETS and the ITS-1 regions leading to 20S pre-rRNA (
; see also , compare Flag-rpS co-purification efficiencies of 20S pre-RNA and 18S rRNA with the ones of 23S and 35S pre-rRNAs at permissive conditions). Hence, most SSU r-proteins show stabilized association with 20S pre-rRNA containing pre-ribosomes. SSU processome dependent pre-rRNA cleavage events leading to 20S pre-rRNA, in particular cleavage at site A2
, were recently suggested to induce a conformational switch in pre-ribosomes 
which might be a prerequisite for distinct r-protein assembly events. Nevertheless, the cleavages leading to 20S pre-rRNA seem not to be sufficient to drive progression of r-protein assembly since tightening of r-protein - pre-rRNA interactions is clearly affected on the level of the residual amounts of 20S pre-rRNA which is still produced in the absence of rpS5 expression 
or after inactivation of Noc4p ().
In contrast to r-proteins, SSU processome components interact strongly with largely un-processed nucleolar pre-rRNPs and weaker with more matured precursor particles (see for example and , compare co-purification efficiencies of 20S pre-RNA with co-purification efficiencies of 23S and 35S pre-rRNAs). Early co-transcriptional and stable binding of SSU processome components could thereby inhibit in vivo
formation of inter- or intramolecular contacts of rRNA precursors which interfere with the establishment of r-protein - pre-rRNA interactions. In agreement with this, the suggested SSU rRNA binding sites of the U3 snoRNA and snR30, another snoRNA essential for early pre-rRNA processing, are incompatible with the two major intramolecular rRNA contacts between the central and 5′ secondary structure domains observed in mature SSUs 
. Enzymatic activities, as for example RNA helicase activities, predicted for a few of the SSU processome components 
, or potential direct contacts between SSU processome sub-modules and r-proteins might also contribute to stabilise transient r-protein-rRNA interactions 
. Future in vitro
studies on the impact of Noc4p and other SSU processome components on pre-rRNA folding and on the assembly of r-proteins should help to understand in more detail the mode of their action in early steps of eukaryotic SSU maturation.
A subset of SSU processome components (Rrp7p/Utp22p group in ), including the RNA helicase Rok1p and the UTP-C sub-module members Rrp7p and Utp22p were identified here to be specifically affected in their association with early SSU precursors after in vivo
depletion of rpS13 and rpS14. The E. coli
homologues of rpS13 and rpS14, S15 and S11, are primary and tertiary binder of one of the central domain assembly trees important for folding of the SSU platform. Inactivation or in vivo
depletion of Rok1p, Rrp7p, Utp22p, rpS13, or rpS14 (and other central domain binders as rpS1 and rpS27) leads to similar early 18S pre-rRNA processing phenotypes 
. Interestingly, overexpression of rpS27, which binds in the SSU rRNA central domain adjacent to rpS13 
, rescues the lethal phenotype of yeast rrp7
deletion mutants 
. In addition, in vivo
depletion of the helicase Rok1p was shown to affect specifically the pre-rRNP association of snR30 
. SnR30 is one of the three small nucleolar RNAs essential for early steps of rRNA maturation 
which was recently shown to bind in vivo
to sequences of the eukaryote specific expansion segment 6 in the rRNA central domain 
. These data further indicate a specific functional link between the SSU central domain assembly state and early SSU precursor interactions of factors as Rok1p and UTP-C sub-module members.
Other SSU processome components (Noc4p/Nop14p group in ) were affected in their association with early SSU precursors not only by in vivo
depletion of rpS13 and of rpS14, but also after shut down of RPS5 expression. RpS5 binds in the SSU head domain adjacent to the platform constituent rpS14. Its E. coli
homologue S7 is the primary binder of the in vitro
assembly tree of SSU head domain r-proteins. Consistent with this, yeast rpS5 is required for efficient in vivo
assembly of the eukaryotic SSU head constituents rpS3, rpS10, rpS15, rpS16, rpS19, rpS20, rpS28 and rpS29 
. Several SSU processome components whose association with early SSU precursors were affected by rpS5 depletion were shown previously to interact with each other or with constituents of the SSU head domain. Interactions between Bms1p and Rcl1p were observed in vitro
and in two hybrid assays 
and were furthermore indicated by ex-vivo
co-purification experiments 
. Large scale analyses revealed genetic interactions between Noc4p and Utp30p 
and between Utp30p and Rrp7p 
. Moreover, Noc4p forms a salt resistant protein complex with Nop14p 
. Nop14p interacts in two hybrid assays with Emg1p/Nep1p 
, a pseudouridine N1-methyltransferase required for methylation of pseudouridine 1191 in the yeast SSU head domain 
. The lethal phenotype of an emg1
deletion mutant strain was shown to be rescued by overexpression of RPS19B 
, whose gene product rpS19 is stably incorporated into the SSU head domain in a Noc4p (see above) and rpS5 dependent way 
. Finally, pre-rRNA interaction sites and localization of Enp1p were recently mapped in the SSU rRNA 3′ domain 
and Enp1-TAP fusion proteins showed reduced efficiency in co-purification of early pre-ribosomal particles after depletion of Noc4p (see Fig. S5
, note that Noc4p depletion did not significantly affect the association of Utp4p, Pwp2p, Utp22p or Imp3p with early pre-ribosomes). In conclusion, these data reinforce the existence of a functional interaction network among members of the Noc4p/Nop14p group () and SSU head domain constituents.
Interestingly, Noc4p was affected in its association with early pre-ribosome by in vivo depletion of rpS5 and the central domain binders rpS13 and rpS14, being itself required for r-protein assembly events in the SSU head domain. One straight forward interpretation of these observations is that a distinct central domain assembly state has to be established to allow efficient recruitment of Noc4p to pre-ribosomes. Noc4p, potentially together with other factors as Nop14p, Emg1p and Enp1p, could then facilitate in a cooperative way downstream r-protein assembly events in the SSU head domain. In such a scenario, the SSU processome component Noc4p coordinates early steps of in vivo folding and assembly of the central and the 3′ major 18S rRNA secondary structure domains thereby providing a quality control checkpoint in the process of eukaryotic SSU assembly.